What we shipped, and why it mattered.
Each entry below is a real release, the date is the ship date, the version is the semver tag. Conservative on claims, specific on commitments. For partners on pilots, the private release notes go out separately under NDA.
Every line item, in order.
Latest release marked in amber, earlier releases in subdued ink. Dates set in tabular mono so the column reads as a ledger.
- v2026.20
Comorbidity-aware grading and a seven-language trial workspace
The grading judge now reasons about pre-existing conditions, the provider roster shows real-time PRO freshness, and the trial workspace ships in seven languages with inline term definitions.
gradingproi18nrelease18 items shippedComorbidity reasoning live7 languages6 audience landingsNew features
Comorbidity-aware adverse event grading
Pre-existing conditions like diabetic neuropathy, COPD, and chronic kidney disease now surface alongside the AE grade, and the judge agent reasons about them when proposing attribution. Coordinators and PIs no longer have to flip back to the chart to weigh a comorbidity against a treatment-emergent grade.
Why it mattersAttribution for patients with complex profiles is where grading variance is highest. Putting comorbidities in front of the grader, and in the model's reasoning, brings the suggested attribution closer to what an experienced PI would write.
Real-time PRO tracking on patient lists
The provider patient roster now shows a per-patient "Last PRO" timestamp directly on the list, and the mobile roster refreshes when you return to the screen. Coordinators can scan a roster, see who has not submitted symptoms recently, and triage from one screen.
Why it mattersFollowing up on missing patient-reported outcomes is one of the daily tasks coordinators repeat most often. Surfacing freshness on the roster removes the drill-down step and shortens triage.
Trial workspace in seven languages, with inline definitions
Terminology tooltips now show inline definitions for over 100 clinical terms (CTCAE grades, SAE reporting, biomarkers, enrollment stages) across the trial workspace. Coverage is live in English, Spanish, French, German, Italian, Chinese, and Japanese.
Why it mattersInternational coordinators onboard faster and the acronym ambiguity that slows protocol reviews drops. The same workspace now serves a global trial network, not just one language.
Audience-specific marketing landings
The public site now has dedicated pages for pharma sponsors, academic medical centers, CROs, community oncology practices, clinicians, and patients, plus engine, validation, and security pages and an AE Labor Cost Calculator. The FAQ hub answers 129 questions across 28 categories.
Why it mattersEach audience now lands on positioning tailored to their workflow instead of a generic homepage. The labor cost calculator gives operators a concrete read on what AE grading time is worth.
Account deletion on mobile
Patient and provider mobile apps now support a unified account deletion experience with clear loading and post-deletion states.
Why it mattersUsers have asked for self-serve account control directly, and it is required by app store guidelines on mobile platforms.
Improvements
- Accurate workflow status on refresh. Workflow tabs now show real in-progress and complete status immediately after a page reload, instead of flashing a stale "idle" state during long grading runs.
- Faster trial workspace in seven languages. Inline definitions and tooltips load without slowing the page.
- Steadier live grading. Live grading now runs through a simpler server-driven pattern. Same experience on screen, fewer moving parts behind it.
This week's impact
- 18 product items merged
- Comorbidity reasoning live in the AE grading judge
- Provider roster now shows real-time PRO freshness
- Trial workspace usable in seven languages with inline term definitions
- Six audience-specific marketing landings published, plus a labor cost calculator
- v2026.19
Protocol ingestion, demo mode, and a mobile-responsive provider portal
Upload a clinical trial protocol PDF and the grading pipeline pulls eligibility criteria and AE specifications directly from it. A new passwordless demo flow gets evaluators into the product in seconds. The provider portal now adapts cleanly to phones and tablets.
gradingmobileauthrelease9 PRs shippedProtocol ingestion with RAG searchPasswordless demo modeMobile-responsive provider portalNew features
Trial protocol ingestion with retrieval-augmented search
Upload a clinical trial protocol PDF and the platform now extracts structured trial metadata automatically: eligibility criteria, adverse event specifications, and dose regimens. The document is embedded for semantic search, and the protocol's own evidence threads into every CTCAE grading run on that trial. Coordinators can review the extracted protocol structure in a dedicated UI before activating it.
Why it mattersSetting up a new trial used to mean re-typing protocol details into the system. Now the protocol is the source of truth and the grading pipeline pulls criteria directly from it, with citations back to the original document.
Demo mode for faster evaluations
A new passwordless demo flow lets prospective customers and partners sign up with name, email, and phone, skip password setup entirely, and land in a working demo organization in seconds. Available at
demo.burna.ai.Why it mattersFewer clicks between "I want to see this" and "I am clicking around the product." Lower friction for sales conversations and partner walkthroughs.
Mobile-responsive provider portal
Home, encounters, and patient pages in the provider app now adapt cleanly to phone and tablet screens. Tables wrap correctly, action buttons collapse into dropdown menus on small screens, headers stack on phones, and spacing tightens on smaller viewports.
Why it mattersCoordinators and clinicians do not always work at a desk. The provider portal is now usable wherever you are, not just on a clinic workstation.
Improvements
- Simpler translation files. The 7-language translation system was consolidated from 22 to 30 JSON files per language down to a single file per language, with zero translation keys lost. Coverage: English, Spanish, French, German, Italian, Chinese, and Japanese.
- Invitation expiry edge cases closed. Member invitation expiry date handling is now backed by a shared date utility with unit test coverage, preventing off-by-one errors.
Bug fixes
- PRO-CTCAE patient symptom reporting log emission and symptom-management edge cases resolved.
- CORS allowlist updated to reflect current deployment environments (
demo.burna.aiadded, deprecatedalpha.burna.airemoved).
Quality and reliability
- A full end-to-end test suite for sign-in, sign-up, and auth redirect now runs on every PR, with a dedicated CI workflow.
- The E2E authentication routes were extracted into a dedicated module so OTP bypass logic in test environments is reliable in preview deployments.
This week's impact
- 9 PRs shipped
- Protocol ingestion with retrieval-augmented search live
- Passwordless demo mode available at
demo.burna.ai - Provider portal fully mobile-responsive
- 7 languages consolidated into single-file format
- End-to-end auth testing live in CI
- v2026.18
Live grading streaming, public website launch, and staging environment
CTCAE grading, drug attribution, and drug interaction results now stream end-to-end in real time. The Burna AI public website is live across all audience landings. PRO enrollment tokens can now be sent directly from the provider app, and a staging environment is available for pre-production validation.
gradinginfrastructuremarketingrelease15 PRs shippedLive workflow streamingPublic website launchedStaging environment liveNew features
Real-time workflow streaming for grading results
Live progress for CTCAE grading, drug attribution, and drug interaction analysis now streams end-to-end through the native workflow protocol instead of being polled from persistent text streams. The provider app sees step status (running, completed, failed) the moment the workflow emits it, with organization context propagated automatically on every request.
Why it mattersWorkflow runs feel instant and consistent. Less waiting, less spinner, and a cleaner foundation for chaining workflows in the future.
Burna AI public website is live
The new Burna AI marketing website is live, covering home, technology, healthcare providers, life sciences, clinical research, patients, careers, press, contact, request-demo, and a trust center and security compliance section. A gated investor portal and a 23-slide deck are available for partner conversations.
Why it mattersOne canonical place for prospects, partners, and investors to learn what Burna AI does, how it works, and how to engage. Replaces the old single-page placeholder.
PRO enrollment token email sent directly from the provider app
The PRO enrollment token email is now wired through the platform end to end. Coordinators can copy a fresh token to the clipboard or send a branded enrollment email to a patient directly from the provider app. Token reissuance is fully audit-logged.
Why it mattersWhen a patient loses their PRO invite, you can get them back into the flow in seconds without leaving the app or filing a support ticket.
Staging environment for QA
The provider and patient apps can now run in a true pre-production environment at
staging-app.burna.aiandstaging-care.burna.ai, with the same authentication and security posture as production.Why it mattersReleases can now be tested against real services before they reach production users.
Improvements
- Faster settings modal. The settings modal loads about 45 KB lighter on first paint, with each tab lazy-loaded only when opened (approximately 400 ms faster Time to Interactive).
- Enterprise admin file access. Knowledge base file access now respects organization admin status independently of member role, unblocking enterprise admins.
- Cleaner filter state. Provider web filter logic refactored to eliminate stale-state issues.
- Dependency updates. React and React DOM bumped to 19.2.5 with the latest TypeScript native preview compiler for faster builds across all 18 packages. Model SDK packages and Anthropic Claude Agent SDK updated for improved streaming token efficiency.
- Smaller bundle. Canvas rendering migrated to platform-specific native binaries for approximately 12% bundle size reduction.
Bug fixes
- Workflow stream step IDs now propagate correctly into the run ID field, so step tracking is no longer lost.
- Authentication mutation hardened with explicit typing and a defensive null check.
- Crisp Chat no longer fails to initialize when a user profile is missing a name field.
- iPad layout regressions and a token login edge case fixed in the patient mobile app.
- CORS credentials behavior in the workflow engine aligned with browser security model.
This week's impact
- 15 PRs shipped
- Live workflow streaming migrated to native protocol
- Public Burna AI marketing website launched
- Staging environment available for pre-production validation
- 2,200+ files changed
- v2026.16
PRO symptom grading, token reissuance, and mobile dark mode
Patient self-reported symptoms now flow through the same gold-standard CTCAE grading pipeline as clinician-entered events. Coordinators can reissue enrollment tokens in one click, and the provider mobile app ships full dark mode, a 50-state address picker, and a graceful workflow timeout.
gradingpromobilerelease9 PRs shippedPRO-to-CTCAE pipeline liveToken reissuance in one clickMobile dark modeNew features
Patient-reported symptoms now auto-graded with the full CTCAE pipeline
When a patient submits a PRO-CTCAE check-in, the grading pipeline now runs in parallel with the existing symptom-triage workflow on the same submission. The system synthesizes a clinical narrative from the structured PRO responses and produces a complete CTCAE grade proposal, cited and ready for clinician sign-off. A rebuilt Symptom Management area in the provider app surfaces the grade queue, detail view with grade definitions, inline editing, and 21 CFR Part 11-compliant e-signature.
Why it mattersPatient self-reports no longer sit in a separate silo. They feed the same gold-standard grading pipeline your team uses for clinician-entered events, with full clinician oversight and cross-organization isolation.
Re-issue PRO enrollment tokens in one click
Coordinators can now re-issue a PRO-CTCAE enrollment token from the provider app when a patient loses access to their original invite. A new dialog invalidates prior tokens and triggers a branded email with the fresh one. No manual edits or support tickets required.
Why it mattersPatients drop off between visits when invites expire or get lost. Re-issuing takes seconds now, not a support escalation.
Provider mobile app: dark mode, state picker, and workflow timeout
The provider mobile app picked up a theme toggle in settings with full dark-mode parity across all screens, a native 50-state picker for patient addresses, and a 5-minute polling timeout on long-running workflows so the UI can guide a retry instead of spinning indefinitely.
Why it mattersMobile clinicians get a polished light and dark experience, faster address data entry, and a graceful failure mode when a workflow stalls on the road.
Improvements
- CTCAE search help dialog. The search page now has a working Help dialog explaining search, results, quick picks, and export.
- Drug attribution column. The patient adverse-events modal now includes a Drug Attribution column that links directly to the encounter.
- Readable workflow output. "View full output" switched from raw JSON to a human-readable key-value tree.
- Resend OTP fully wired. The signup "Send Another Link" button now has loading state and toast feedback.
- Active patient badge. The patient list shows a green "Active" badge instead of a dash placeholder.
- Confidence display. AI confidence is now shown as a color-coded badge with a percentage.
- Patient mobile: no active trial gate. The diary now renders a clear enrollment prompt instead of a generic error when no trial is active.
- Patient mobile: greeting by name. The home screen greets patients by their linked EHR profile first name.
- Safe area handling. The Add Diary Entry screen uses consistent top spacing across notched and non-notched devices.
- Navigation fixes. Help navigation on the patient diary stops stacking duplicate routes; the provider home patient list scrolls to top when switching alphabet letters.
- Bottom sheet discoverability. Add-Patient bottom sheet footer actions are always visible.
Bug fixes
- Gender field now propagates correctly through patient creation flows and normalizes to the backend enum.
- Patient queries no longer fire before authentication is ready.
- Drug attribution shows the actual Kramer category instead of a hardcoded "definite."
- PDF downloads now use a reliable cross-origin delivery method.
- Symptom management baseline mapping now includes all required fields.
- SAE reporting weight label corrected to "Weight (lbs)."
- CTCAE search no longer flashes "no results" while the user is still typing.
- Rationale text preserves line breaks instead of collapsing them.
This week's impact
- 9 PRs shipped
- Patient-reported symptoms now feed the full CTCAE grading pipeline
- PRO enrollment token reissuance live in the provider app
- Provider mobile app ships dark mode, state picker, and workflow timeout
- 110+ files changed
- v2026.15
Knowledge-powered grading, patient mobile app, and organization-level EHR management
Upload your protocols and clinical guidelines for AI grading to reference during assessments. The patient mobile app launches with symptom diary and safety reporting. EHR connections are now managed at the organization level for the whole team at once.
gradingmobilesafetyrelease11 PRs shippedPatient mobile app live5,800+ Epic endpoints1,900+ Cerner endpointsNew features
Knowledge-powered AI grading
Upload your organization's protocols, drug labels, and clinical guidelines, and the AI grading system will automatically reference them during CTCAE assessments. The system parses, indexes, and retrieves relevant context from your documents in real time. If no documents are uploaded, grading continues as normal with zero disruption.
Why it mattersYour AI grading is now informed by the same institutional knowledge your clinicians use, producing more accurate and contextually relevant CTCAE grades for your specific protocols and patient population.
Protocol management for multi-drug attribution
Clinical coordinators can now manually enter expected adverse event profiles for trial protocols, with CTCAE term auto-completion. These profiles are automatically injected into multi-drug attribution analysis, anchoring assessments to known drug-event associations. Connect to Protocol Safe for automated sponsor-hosted protocol data delivery. Full audit logging is included for regulatory compliance.
Why it mattersAttribution analysis is now grounded in protocol-specific expectations, reducing assessment time for combination therapy cases where drug-event associations are complex.
Patient mobile app
A complete patient-facing mobile experience is now available. Patients can authenticate with email or phone via one-time passcode, enroll in a trial via token, access a home screen with quick actions, maintain a symptom diary with expandable entries, and submit safety reports. The app supports dark and light mode across all screens.
Why it mattersPatients can now self-report symptoms between clinic visits directly from their phone, enabling continuous monitoring and earlier intervention for treatment-emergent adverse events.
Organization-level EHR management
EHR base URLs for Epic, Cerner, and Athena are now managed at the organization level instead of per user. When an admin updates an endpoint, all team members' connections are automatically refreshed. A searchable endpoint selector covers 5,800+ Epic and 1,900+ Cerner endpoints.
Why it mattersOne control point manages your entire team's EHR connectivity. Admins no longer need to troubleshoot individual user configurations when endpoints change.
Dynamic EHR OAuth discovery
OAuth token endpoints are now resolved dynamically using SMART configuration endpoints instead of hardcoded values. The system automatically adapts when EHR vendors update their infrastructure.
Why it mattersEHR integrations stay connected even when providers change their infrastructure, eliminating silent token refresh failures that previously required manual intervention.
Improvements
- Infinite scroll on mobile. Cases and patient lists on mobile now use infinite scroll pagination for faster browsing of large lists.
- MRN field on patient forms. MRN can now be added or edited directly on the patient create and edit forms.
- FAQ search on mobile. The mobile help screen now supports FAQ search filtering.
- More accurate review status. Cases list now uses client-side filtering for more accurate review status display.
- Unified patient edit form. Patient edit experience is now consistent across the bottom sheet and standalone views.
- Reduced redundant API calls. Workflow data is persisted so repeated calls to the workflow engine for the same data are eliminated.
Bug fixes
- Fixed grading data extraction path to match the web workflow structure.
- Fixed a logout race condition that caused authentication errors on session end.
- Fixed login button contrast for better visibility.
- Fixed FAQ text spacing issues on mobile.
- Fixed the recording countdown timer.
This week's impact
- 11 PRs shipped
- 3 major platform features launched
- 150+ files changed
- 2 contributors
- v2026.14
Clinical preferences, expanded EHR data access, and severity-first AE review
Organizations can now set institution-specific rules that guide AI grading across all workflow categories. Cerner integration expands to 110+ FHIR resource types, and adverse events sort by severity so the most critical cases surface first.
gradingsafetycompliancerelease13 PRs shipped5 new features110+ FHIR resource types14 new integration testsNew features
Clinical preferences for AI grading
Organizations can now configure institution-specific rules that guide AI grading across all workflow categories: CTCAE grading, drug attribution, drug interaction analysis, and encounter title generation. Set once in Settings, and every AI suggestion follows your clinical protocols.
Why it mattersDifferent institutions have different grading conventions and clinical priorities. Clinical Preferences ensure AI suggestions align with how your team practices, reducing review time and improving consistency across your organization.
Expanded EHR integration
Cerner/Oracle Health integration now supports 110+ FHIR resource types, up from 40. Coverage now includes Encounters, Observations, Procedures, Medications, CarePlans, Conditions, MedicationRequests, and DiagnosticReports. OAuth error handling has been improved across all EHR connections with better redirect flows when authentication fails.
Why it mattersBroader data access provides richer clinical context for AI grading. Conditions, medication requests, and diagnostic reports feed directly into adverse event analysis, improving grading accuracy for patients on complex regimens.
Connect to EHR from the home screen
Users without an EHR connection now see a prominent "Connect to EHR" option directly on the clinical note input. One-click OAuth linking for Epic, Oracle Health (Cerner), and AthenaHealth is available right where it is needed.
Why it mattersNew users no longer need to navigate to Settings to connect their EHR. The prompt appears where it is most relevant, reducing setup friction.
Severity-based adverse event ordering
Adverse events now sort by grade, with Grade 5 at the top and Grade 1 at the bottom. High-severity events (Grades 3 through 5) are expanded by default; lower-severity events are collapsed to reduce visual noise during clinical review.
Why it mattersCritical adverse events surface immediately. Coordinators reviewing a patient's AE list can focus on what matters most without scrolling through lower-grade events first.
Grade descriptor tooltips
Hovering over any CTCAE grade badge now shows the grade descriptor text, providing quick clinical context without opening the full definitions dialog.
Why it mattersCoordinators can confirm grade criteria at a glance without interrupting their review workflow.
Improvements
- CSV export for attribution and interaction history. Drug attribution and drug interaction history can now be exported as CSV files, in addition to the existing JSON export.
- Soft-delete for clinical documentation. Clinical records are now archived with timestamps rather than hard-deleted, preserving audit history.
- Slack integration improvements. Added a connectivity health check and richer support ticket notifications.
- Improved invitation flow. Wrong-account errors during organization invitations are now handled gracefully with an automatic redirect.
- Archived record filtering. Visit summary queries now correctly exclude archived records from active views.
- Type safety improvements. More than 20 type safety improvements across data deletion workflows.
Bug fixes
- Fixed the organization invitation flow failing when the user is signed into the wrong account.
- Fixed archived visit summaries appearing in active queries.
This week's impact
- 13 PRs shipped
- 5 new features deployed
- 95+ files changed
- 14 new integration tests added
- v2026.13
Regulatory report generation, study site management, and visit documentation
Generate FDA MedWatch, IRB, and Sponsor SAE reports directly from adverse event data. Study sites and trial protocols are now managed in one place, and clinical visits receive AI-generated titles and summaries automatically.
safetycompliancegradingrelease9 PRs shipped3 regulatory report typesStudy site management liveAuto-generated visit titlesNew features
Regulatory report generation
Generate FDA MedWatch 3500A, IRB Unanticipated Problem, and Sponsor SAE reports directly from adverse event data. Each report type includes compliance tracking, validation rules, and supplemental data collection so regulatory submissions are consistent and audit-ready.
Why it mattersRegulatory reporting is one of the most time-consuming and error-prone parts of clinical trial management. Automated report generation with built-in validation reduces manual effort and helps ensure nothing is missed before submission.
Automated visit titles and excerpts
Clinical notes now automatically receive AI-generated titles and summaries. When a visit with clinical notes is created, the system produces a concise, clinically appropriate title and excerpt, so your visit list is immediately useful without manual documentation.
Why it mattersResearch coordinators no longer need to spend time writing summaries for each visit. The AI handles it in seconds, and the results appear automatically in the visit record.
Study site management
Create, search, view, and edit study sites with full institutional details including location, contact information, principal investigator, and coordinator. Study site data is organization-scoped for multi-site trials.
Why it mattersStudy site data is foundational for regulatory reporting and patient-facing trial information. Managing it in one place eliminates scattered spreadsheets and ensures accuracy across all reports.
Protocols and PRO enrollment administration
New admin pages support managing clinical trial protocols with IND and IRB numbers, sponsor details, and regimen data. Coordinators can issue passwordless PRO portal enrollment tokens tied to specific protocols, sites, and cycles.
Why it mattersTrial setup and patient enrollment workflows are more streamlined, reducing administrative overhead for research coordinators.
Organization-wide visit sharing
Visits can now be shared across the entire workspace, giving team members visibility into organization-wide clinical activity.
Why it mattersCollaboration across sites and teams improves because visit data is accessible to authorized users without manual sharing steps.
Improvements
- Upgraded AI grading model. The grading and clinical analysis engine has been upgraded for more accurate CTCAE assessments.
- Improved billing flexibility. Migrated to a plan-based billing architecture for more flexible subscription management.
- Next.js upgraded to 16.2.1. Latest framework improvements for performance and stability.
- Consistent alert colors. Alert components now use dedicated color tokens for consistent appearance in light and dark mode.
- Cleaner code patterns. Optional chaining standardized across the platform for consistency and reliability.
Bug fixes
- Fixed blind grading visibility for organization admins who are also assigned as raters. Rater role now correctly takes priority.
- Fixed dark mode color issues across signup, settings, and search pages.
- Fixed broken integration logos for Slack, Gmail, Google Calendar, and Google Drive.
- Fixed microphone permission handling: mic access is now deferred until recording starts, removing an unnecessary loading spinner on page load.
- Fixed stream handling in workflow content for proper cleanup and resource management.
This week's impact
- 9 PRs shipped
- 430+ files changed
- 3 regulatory report types added
- AI workflows now auto-generate visit documentation
- Full study site and protocol management launched
- v2026.12
Clinical AE reports, parallel workflow orchestration, and real-time streaming
Adverse event reports are now audit-ready PDFs with full attribution and regulatory flags. The platform now runs CTCAE grading, drug attribution, and interaction analysis in one pass, with each step streaming progress independently.
gradingsafetyreliabilityrelease12 PRs shippedAE reports liveParallel workflow orchestrationReal-time step streamingNew features
Clinical-grade AE report PDF
Full branded adverse event reports are now available as a single print-ready document. Each report includes severity-colored cards, drug attribution with WHO-UMC and Kramer causality categories, regulatory cascade flags for FDA MedWatch, IRB, and Sponsor SAE submissions, patient context, study site data, and summary statistics.
Why it mattersClinical teams can generate audit-ready adverse event reports with complete attribution and regulatory flagging in seconds, replacing a manual assembly process that previously took hours.
Parallel workflow orchestration
When CTCAE grading completes, drug attribution and drug interaction analysis now start automatically and run in parallel. Clinicians see all three analyses side-by-side with real-time streaming progress for each step.
Why it mattersThe manual step of separately triggering attribution and interaction analyses after grading is gone. Clinicians get a complete adverse event picture in one pass, not three.
Real-time workflow streaming
Each workflow step, from term extraction and drug standardization through batch lookup, grading judge, and output assembly, now streams progress independently. Clinicians can review early results while later steps are still processing.
Why it mattersPerceived wait time during AI-assisted grading drops significantly. Early results are available to review before the full workflow completes.
Alpha environment for early access testing
A dedicated early-access environment is now available for design partners and early testers to validate new features before production rollout. Authentication flows work seamlessly across all environments.
Why it mattersDesign partners can validate features in a dedicated environment without touching production workflows.
Improvements
- Clearer clinical labeling. "Drug Interaction" sections renamed to "Clinical Evidence" for improved clinical accuracy across web and mobile. "Suggested" labels updated to "AI Suggested" to clearly distinguish AI proposals from human grades.
- Redesigned AE report cards. Color-coded severity indicators and a streamlined card layout make severity clear at a glance.
- Patient assignment required before recording. Starting a recording session now requires a patient to be assigned first, preventing orphaned recordings.
- Improved breadcrumb navigation. Cleaner navigation across encounter, grading, and analytics pages.
Bug fixes
- OTP sign-in improved: expiry extended from 10 to 15 minutes, session now persists across page reloads, and error messages are clearer.
- Patient MRN field now saves correctly across all creation workflows.
- CTCAE term search results render correctly after a parameter alignment fix with the lookup tool.
- Recording flow now shows a proper "Starting..." state between countdown and actual recording start.
This week's impact
- 12 PRs shipped
- 200+ files changed
- 3 major workflow features deployed
- 6 bug fixes improving daily clinical workflows
- v2026.11
EHR encounter import, batch adjudication, and expanded CTCAE terminology
Providers can now pull clinical notes directly from Epic and Cerner into the grading workflow, resolve multiple grade disagreements in a single submission, and benefit from 150+ CTCAE synonym terms covering drug-specific toxicity and immunotherapy patterns.
gradingsafetyreliabilityrelease7 PRs shippedEHR encounter import150+ synonym terms40% faster adjudicationNew features
Import encounters directly from your EHR
Pull clinical notes straight from Epic or Cerner into the CTCAE grading workflow. Search for patients by name, date of birth, or ID; browse their encounters by date range; and import notes with a single click. Patient assignment happens automatically, and you can undo imports if needed.
Why it mattersNo more copy-pasting notes between systems. Clinical documentation flows directly from your EHR into grading, saving time and reducing transcription errors.
Manual report retry
If automatic report generation fails, you can now retry directly from the interface. The system checks for existing reports to avoid duplicates, and a built-in grace period prevents premature retries.
Why it mattersYou stay in control of report generation without needing to contact support or wait for a background process to catch up.
Improved adjudication workflow
The adjudication interface for resolving grade disagreements has been redesigned. Multiple disagreements can now be resolved in a single submission instead of one at a time. The new batch resolution is approximately 40% faster.
Why it mattersAdjudicators can resolve cases more efficiently, keeping grading workflows moving and reducing bottleneck delays during multi-rater reviews.
Updated result displays
New collapsible sections for CTCAE results and drug attribution data make it easier to review grading details without information overload. Drug interaction attribution is now displayed more clearly.
Why it mattersCleaner presentation helps clinicians focus on the most relevant grading information during review.
Improvements
- Broader CTCAE recognition. Synonym coverage expanded from 73 to 150+ terms, now including drug-specific toxicity language, immune-related adverse event (irAE) terminology, and CAR-T/checkpoint inhibitor patterns.
- Clearer clinical naming. "Drug Interaction" workflow replaces the previous "Clinical Decision" label for more precise clinical semantics.
- Streamlined workflow architecture. Centralized trigger handling across recording-based and text-based visits reduces redundancy and improves reliability.
- EHR sandbox mode. Clinical notes are always available during development and testing, keeping the development loop fast.
Bug fixes
- Fixed SAE count display to always reflect the current graded events, ensuring regulatory accuracy.
- Resolved report generation failures caused by workflow step naming mismatches.
- Fixed PDF generation crash with special characters (Unicode arrows, curly quotes).
- Corrected clinical note data extraction from EHR when notes are embedded as base64 data.
This week's impact
- 7 PRs shipped across 736+ files
- CTCAE synonym terms expanded from 73 to 150+
- Adjudication resolution approximately 40% faster with batch submissions
- Direct EHR encounter import from Epic and Cerner
- v2026.11b
Comorbidity-aware grading, copy-forward detection, and 42 regimen profiles
Burna AI now factors patient comorbidities into CTCAE grading, detects stale copy-forward documentation, learns new clinical vocabulary from unmatched terms, and ships pre-built profiles for 42 oncology combination regimens.
gradingsafetypharmacovigilancerelease4 clinical intelligence features42 regimen profiles458+ terminology terms5 new organ system areasNew features
Comorbidity-aware grading
Burna AI now considers patient comorbidities during CTCAE grading. When a patient with Type 2 Diabetes presents with peripheral neuropathy, the system recognizes this may be a pre-existing condition rather than a new treatment-related adverse event. This reduces false-positive AE attribution and saves coordinators the 5 to 10 minutes they currently spend manually cross-referencing comorbidities per patient.
Why it mattersMore accurate grades with less manual work. The system flags when an adverse event may overlap with a known condition, giving clinicians the context they need to make confident decisions.
Copy-forward detection
Clinical documentation often carries forward unchanged notes from prior visits. Burna AI now detects when adverse event descriptions and grades remain identical across multiple visits, flagging potential documentation artifacts for clinician review. This helps ensure grades reflect a patient's current condition, not stale data from weeks or months ago.
Why it mattersStale grades can delay dose modifications and create regulatory risk during audits. This feature catches what manual review often misses.
Self-improving terminology
When the system encounters medical terms it has not seen before, it no longer silently skips them. Unmatched terms are captured and routed to a clinician-guided review queue. Once resolved, the system learns the new vocabulary permanently. Every clinical note processed makes the system smarter.
Why it mattersClinical vocabulary varies across institutions. A system that adapts to your team's language gets more accurate over time, not less.
42 oncology regimen profiles
Pre-built drug profiles for the most common oncology combination regimens, including FOLFOX, FOLFIRI, carboplatin/paclitaxel, and checkpoint inhibitor combinations. Multi-drug attribution works out of the box without custom configuration.
Why it mattersWhen a patient is on a 3-drug combination regimen, attributing an adverse event to the right drug matters for dose decisions. These profiles encode the known toxicity patterns for 42 regimens so attribution is immediate.
Improvements
- Broader terminology coverage. Medical terminology now spans 26 organ system classes with 79 groupings covering 458+ individual terms, including new coverage for hepatobiliary, ear/labyrinth, and reproductive system toxicities.
- Faster grading pipeline. Streamlined internal workflow architecture improves overall grading performance.
- Improved system reliability. Consolidated configuration layer reduces surface area for configuration drift.
This week's impact
- 4 major clinical intelligence features shipped
- Coverage expanded across 5 new organ system areas
- 42 regimen profiles available out of the box
- System now learns and improves with every clinical note processed
- v2026.10
EHR patient import, AE report PDFs, and provider safety alerts
Coordinators can now import patients directly from Epic, Cerner, and Athena, generate audit-ready adverse event report PDFs in seconds, and receive automatic alerts when patients self-report Grade 3 or higher events.
gradingsafetyinfrastructurerelease8 feature PRs shippedEHR import liveAE report PDFsGrade 3+ provider alertsNew features
Import patients directly from your EHR
Search and import patient records from Epic, Cerner, and Athena without leaving the app. When launched via SMART on FHIR, the patient context loads automatically with a one-click import. In standalone mode, search by name, date of birth, gender, or MRN. Imported demographics stay in sync, and EHR-sourced fields like name and MRN are protected from accidental edits.
Why it mattersEliminates dual data entry between your EHR and Burna AI. Patient records flow directly from your existing systems, reducing errors and saving coordinators time on every new encounter.
Workflow resumption for failed steps
If a network hiccup interrupts an attribution or clinical decision workflow mid-run, clinicians can now retry the failed step without re-doing the entire CTCAE grading. A "Retry Workflow" button appears in the error state and picks up where the workflow left off.
Why it mattersIn busy clinical settings, transient failures are inevitable. Resumption capability means no lost work and no repeated effort, keeping coordinators moving through their case queue.
AE report PDF generation
Generate structured adverse event report PDFs directly from the platform. Reports include all grading data, citations, and audit trail information in a format suitable for regulatory submissions and sponsor reporting.
Why it mattersCoordinators can produce audit-ready documentation in seconds instead of manually assembling reports from multiple sources.
Provider notifications for high-grade events
When a patient self-reports a Grade 3 or higher adverse event via PRO-CTCAE, the assigned provider receives an automatic notification with a time-based action deadline.
Why it mattersHigh-grade adverse events demand urgent attention. Automated alerts ensure no critical patient safety signal goes unnoticed between clinic visits.
Operational automation hub
52 automation workflows covering customer support, content generation, SEO intelligence, and sales operations. All workflows include error handling, retry policies, and audit logging.
Why it mattersOperational automation frees the team to focus on product development and clinical partnerships rather than manual processes.
Improvements
- More secure EHR access. EHR OAuth flow upgraded from system-level to clinician-level scopes for more appropriate, per-user data access.
- Consistent navigation state. Visit context is now shared across all dashboard routes, so navigating between pages no longer drops your current session state.
- Cleaner audit trails. Workflow data architecture is decoupled from visit records, enabling cleaner audit trails and future workflow types.
- Flexible check-in scheduling. Check-in frequency options expanded to include biweekly, monthly, and adaptive scheduling.
Bug fixes
- Fixed clinical note input staying disabled during SMART on FHIR patient assignment.
- Fixed OAuth token refresh failing due to network access restrictions.
- Fixed URL mismatch errors caused by trailing slashes in OAuth redirect URIs.
This week's impact
- 8 feature PRs shipped
- 5 improvements deployed
- 5 bugs resolved
- 1,000+ files updated across the platform
- v2026.09
EHR integration live, submission-gated blinding, and organization management
Connect directly to Epic, Oracle Health, or Athenahealth through standard clinical authentication protocols. Grading blinding is now enforced at the submission boundary, and a new organization picker streamlines multi-site team management.
gradingauthcompliancerelease9 features shipped5 improvements deployed5 bugs fixed160+ files changedNew features
EHR integration goes live
Connect directly to your Epic, Oracle Health (Cerner), or Athenahealth EHR system with a single click. The platform now supports clinician-level OAuth authentication through standard SMART on FHIR protocols, so your team can securely access patient data without additional IT lift.
Why it mattersEliminating manual data re-entry between your EHR and the grading workflow saves coordinators time on every adverse event assessment and reduces the risk of transcription errors that affect data quality.
Submission-gated blinding for grading integrity
The grading workflow now enforces strict blinding rules: admins and adjudicators can only view a rater's work after that rater has formally submitted their grades. Progress indicators show real-time status (in-progress vs. submitted) without exposing intermediate work.
Why it mattersGrading integrity depends on independent assessment. Submission-gated blinding preserves that independence while giving administrators clear visibility into team progress and completion status, supporting audit readiness for blind grading programs.
Organization management made simple
A dedicated organization picker page streamlines onboarding and team switching. Create new organizations, switch between existing ones, and manage team settings from a single interface.
Why it mattersMulti-site teams can now manage their organizations without navigating complex settings menus, reducing onboarding friction for new team members joining an active trial.
MRN-powered patient search
Search across patient names and Medical Record Numbers simultaneously. The composite search field updates automatically when patient records change, so results are always current.
Why it mattersCoordinators can locate patients instantly using the identifier most natural to their workflow, whether that is a name or MRN, without switching search modes or navigating separate lookup tools.
Real-time analytics dashboard
The analytics dashboard now displays live production data instead of placeholder metrics. Four dedicated tabs (Overview, Productivity, Quality, Practice-wide) show grading volume, agreement rates, turnaround times, and feature adoption across your organization.
Why it mattersSite leadership and coordinators get real-time visibility into grading performance, enabling data-driven decisions about resource allocation and workflow optimization without waiting for a weekly export.
Improvements
- Organization-scoped access enforcement ensures graders only see cases within their assigned organization.
- Adjudicator flow now streams grading events and case status changes to analytics in real time.
- Settings modal stays open after saving, reducing unnecessary navigation.
- Patient creation now automatically assigns new patients to the active visit.
- Integrations page uses reactive data patterns for faster, more reliable EHR status updates.
Bug fixes
- Fixed login redirect sending users to a non-existent page after email verification.
- Fixed logout redirect pointing to the incorrect sign-in URL.
- Resolved auth layout scrolling issues on smaller screens.
This week's impact
- 9 features shipped
- 5 improvements deployed
- 5 bugs fixed
- 160+ files changed across the platform
- v2026.08
Automated adjudication, real-time analytics, and searchable CTCAE terms
Grade conflicts now route automatically to pre-designated adjudicators, the analytics dashboard shows live production data across four tabs, and over 1,000 CTCAE terms are searchable and grouped by System Organ Class.
gradingsafetypharmacovigilancerelease14 PRs merged9 features shippedLive analytics across 4 tabs1,000+ searchable CTCAE termsNew features
Automated adjudicator workflow with collision detection
When two independent raters disagree on a CTCAE grade, the system now automatically detects the conflict and assigns an adjudicator from a pre-designated pool. Adjudicators see both grades side by side and resolve with one click. Built-in collision detection prevents duplicate assignments, and digest notifications keep the team informed without inbox overload. Every resolution includes e-signature support for regulatory compliance.
Why it mattersGrade disagreements are inevitable in multi-rater workflows. Automated adjudication removes manual coordination overhead and ensures conflicts get resolved consistently. For sites running blind grading programs, this is critical: disagreements must be documented and resolved for audit readiness.
Adjudicator role management
Coordinators can now pre-designate team members as adjudicators through a dedicated management interface. When grading conflicts arise, the system automatically assigns from this pool rather than requiring manual selection.
Why it mattersPre-designating adjudicators before conflicts occur means faster resolution when disagreements happen. No more ad-hoc coordination to find someone qualified to break a tie.
Real-time analytics dashboard
The analytics dashboard now shows live production metrics. Four tabs (Overview, Productivity, Quality, Practice-wide) stream live data from clinical workflows, covering active cases, grading volume, agreement rates, and turnaround metrics.
Why it mattersClinical research directors need visibility into grading operations without exporting data to spreadsheets. Real-time analytics provide instant answers to questions like "How many cases were graded this week?" and "What is our inter-rater agreement rate?"
Smart encounter capture modal
New modal-driven encounter creation with smart patient preset. When viewing a patient's profile, starting a new encounter pre-populates patient information automatically. Keyboard accessibility is preserved throughout.
Why it mattersReducing clicks and data entry per encounter adds up. Pre-populating patient information from context eliminates a common source of data entry errors and saves time on every encounter.
Searchable CTCAE term lookup
Over 1,000 CTCAE terms are now searchable and grouped by System Organ Class. Selecting a term auto-populates the SOC field, replacing the previous two-step dropdown that required navigating SOC first, then finding the specific term.
Why it mattersFaster term selection means less friction during manual grading entry. Grouping by SOC helps clinicians find related terms and ensures correct classification without memorizing the full hierarchy.
Manual grading entry
Clinicians can now manually enter CTCAE grades through a dedicated interface when the assisted grading workflow is not needed or available. Supports all grade levels (1 through 5) with term search and SOC auto-population.
Why it mattersNot every adverse event requires the assisted workflow. A clean manual entry path ensures the platform works for all grading scenarios, including retrospective data entry and cases where the assisted workflow does not apply.
Start visit from patient view
Clinicians can initiate a new encounter directly from a patient's profile page. The visit creation flow carries forward patient identifiers and reduces navigation steps.
Why it mattersClinicians think patient first, then encounter. The interface now matches that mental model, reducing the navigation steps between context and action.
Improvements
- Adjudicator collision detection prevents duplicate case assignments during concurrent operations.
- Digest notifications for adjudication events reduce email noise while keeping teams informed.
- CTCAE term search groups 1,000+ terms by System Organ Class with auto-population.
- "Clinical Evidence" labels standardized to "Drug Interaction" across web and mobile platforms.
- Skeleton loaders replaced with consistent loading indicators.
- Completion checkmarks and collapsible causality details added to workflow tabs.
Removed
Treatment recommendations removed from multi-drug attribution. Clinical review determined that generating treatment suggestions without full patient context (comorbidities, contraindications, concurrent medications) could be misleading. Grades and causality remain; treatment decisions stay with the clinician.
Bug fixes
- Fixed patient list item color contrast for better readability in light and dark mode.
- Fixed review status not updating correctly after grading submission.
This week's impact
- 14 PRs merged across the platform
- 9 features shipped: adjudicator workflow, role management, analytics dashboard, encounter capture, CTCAE search, manual grading, patient view visits, infrastructure, and content
- Automated adjudication replaces manual conflict resolution workflows
- Real-time analytics replace all mock dashboard data
- Responsible product decision: removed a feature that lacked sufficient patient context
- v2026.07
Blind grading, grader roles, and a redesigned marketing site
Independent graders can now evaluate cases without seeing prior assessments, coordinators can assign cases in bulk, and clinicians get instant access to all five CTCAE grade descriptors inline.
gradingcompliancemarketingrelease14 PRs merged6 major features shipped400+ files changedFull dark mode on marketing siteNew features
Blind grading for quality assurance
Independent graders can now evaluate CTCAE cases without seeing previous assessments, ensuring unbiased quality control. Coordinators can seed standardized test cases (15 realistic clinical scenarios) and assign graders in bulk, making inter-rater reliability validation and grader certification programs possible.
Why it mattersBlind grading is a cornerstone of clinical trial quality assurance. This feature enables sites to validate grading accuracy, measure consistency across raters, and meet regulatory expectations for AI-assisted grading tool validation in trials.
Grader role and bulk assignment
A dedicated grader role streamlines case routing. Coordinators can filter for qualified raters, assign 2 to 10 cases at once, and see rater names directly in case views.
Why it mattersEfficient case assignment reduces administrative overhead and keeps grading workflows moving, especially during high-volume trial periods when one-by-one assignment is impractical.
Grade definitions at your fingertips
Clinicians can now click any adverse event's "Grade Definitions" button to view all five CTCAE v6.0 grade descriptors, with the currently assigned grade visually highlighted.
Why it mattersQuick access to grade definitions during review improves accuracy and confidence in grading decisions, without switching between documents or looking up criteria separately.
Patient assignment across web and mobile
Assign patients directly within the CTCAE grading workflow on both web and mobile. The web app holds note submission until a patient is assigned, while the mobile app offers a dedicated patients screen with search, filtering, and inline editing. A five-second countdown before voice recording gives clear feedback and recovery time.
Why it mattersRemoving friction from the patient assignment step keeps clinicians focused on documentation rather than navigating between screens, and the countdown provides a moment to confirm before committing to a recording.
Redesigned marketing website
Brand new home page with comparison tables, trust badges (SOC 2, HIPAA, 21 CFR Part 11), demo video, and waitlist form. Six new pages added: About Us, FAQs, Cookie Policy, Privacy Policy, Security and Compliance, and Terms and Conditions, all with full dark mode support.
Why it mattersA professional, transparent web presence builds trust with clinical research leaders evaluating new technology partners. Clear compliance and security pages answer the questions that enterprise procurement teams ask first.
Improvements
- Marketing design system migrated to semantic primary color palette with automatic dark mode support (approximately 280 color updates).
- Voice recording now includes error recovery with retry and manual fallback options.
- Recording state properly resets between patient sessions, preventing stale data.
- Encounter screen centered on wide displays with clearer attribution for accepted records.
- Patient form keyboard navigation improved on mobile for smooth field-to-field flow.
- Workflow tooling migrated from experimental to stable APIs across approximately 20 tools.
Bug fixes
- Fixed encounter content alignment on wide screens.
- Fixed mobile bottom sheet layout with predictable snap points.
- Fixed keyboard dismissal after clipboard paste in recording view.
This week's impact
- 14 PRs merged across the platform
- 6 major features shipped: blind grading, grader roles, grade definitions, patient assignment, marketing redesign, and voice coaching
- 400+ files changed
- Full dark mode across the marketing site
- Blind grading infrastructure enables quality assurance programs
- v0.12
Patient safety dashboard and encounter review improvements
Faster patient safety assessment and better review workflows.
safetygradingmobilereleaseConsolidated AE historyAI-assisted review workflowEncounter audio playback18 features shippedNew features
Patient safety dashboard: view all adverse events in one place
Clinical coordinators can now view a patient's complete adverse event history across all encounters in a single modal. No more navigating through individual encounters to assess safety trends. Includes grade badges, review status, and direct links to source encounters.
Why it mattersReduces the time to assess patient safety from navigating 10+ encounters to viewing one consolidated view.
Accept, edit, and add adverse events during review
Reviewers can now accept AI-suggested adverse events, edit grades and evidence, or manually add new adverse events directly from the encounter review interface. All changes are tracked with accept/edit status for audit purposes.
Why it mattersClinicians maintain full control over AI suggestions with a streamlined review workflow that persists every decision.
Patient assignment across all workflows
Patient assignment is now available everywhere you need it: clinical note input, encounter details, and new encounter modals. The redesigned workflow decouples patient selection from note creation, enabling flexible assignment at any point in the encounter.
Why it mattersFaster encounter setup and better data organization, especially for multi-patient clinical trials.
Unified CTCAE workflow across web and mobile
The CTCAE grading workflow now behaves consistently on both web and mobile platforms. Shared types, constants, and workflow infrastructure ensure the same adverse event data flows through both platforms identically.
Why it mattersWhether you are at your desk or on the ward with a mobile device, the grading experience and data quality are the same.
Encounter audio playback
Listen to encounter recordings directly from the encounter detail view with secure, time-limited audio URLs. The new audio player supports playback speed control and is fully accessible.
Why it mattersReview the original clinical conversation alongside AI-generated notes without switching tools.
Real-time review status tracking
Encounter review status (Pending, Accepted, Rejected) now updates automatically whenever a review is saved. No manual status management needed.
Why it mattersAlways know which encounters need attention. The system keeps status in sync across all views automatically.
Improvements
- Smarter encounter list with paginated loading, collapsible adverse event rows, and grade badges (G1-G5 with severity colors)
- Patient encounters sidebar for quick browsing of all encounters for a specific patient
- Clinical note panel now shows recent encounters for the same patient (last 30 encounters)
- Design system enforcement across 370+ files: all colors now use semantic tokens for consistent light/dark mode
- Mobile workflow error recovery: retry failed recordings without losing completed steps
Bug fixes
- Improved audio player accessibility with proper ARIA attributes
- Fixed navigation path from
/encounter/to/encounters/ - Fixed invalid component prop types
- v2026.06
Complete adverse event history, audio playback, and real-time review status
Coordinators can now view a patient's full adverse event history in one place, listen to encounter recordings inline, and watch review status update automatically across all views.
gradingsafetyreliabilityrelease18 items shipped7 bug fixes4,000+ lines of legacy code removedNew features
Patient safety dashboard: full adverse event history in one place
Clinical coordinators can now view a patient's complete adverse event history across all encounters in a single modal. Grade badges, review status, and direct links to source encounters are all visible without navigating through individual records.
Why it mattersReducing the time to assess patient safety from navigating 10 or more encounters to viewing one consolidated view is a meaningful daily time saving for coordinators managing complex, multi-encounter patients.
Accept, edit, and add adverse events during review
Reviewers can now accept suggested adverse events, edit grades and evidence, or manually add new adverse events directly from the encounter review interface. All changes are tracked with accept or edit status for audit purposes.
Why it mattersClinicians maintain full control over every suggested grade through a streamlined review workflow that persists every decision, satisfying audit-readiness requirements without extra steps.
Patient assignment across all workflows
Patient assignment is now available everywhere you need it: clinical note input, encounter details, and new encounter modals. The redesigned workflow decouples patient selection from note creation, enabling flexible assignment at any point.
Why it mattersFaster encounter setup and better data organization, especially for multi-patient clinical trials where coordinators often start a note before confirming which patient it belongs to.
Unified CTCAE workflow across web and mobile
The CTCAE grading workflow now behaves consistently on both web and mobile platforms. The same adverse event data flows through both platforms identically.
Why it mattersWhether you are at your desk or on the ward with a mobile device, the grading experience and data quality are the same. Coordinators can start on one surface and continue on another without losing context.
Encounter audio playback
Listen to encounter recordings directly from the encounter detail view with secure, time-limited audio URLs. The audio player supports playback speed control and is fully accessible.
Why it mattersReviewing the original clinical conversation alongside notes removes the need to switch tools, cutting the time to verify a disputed detail or clarify an ambiguous term.
Real-time review status tracking
Encounter review status (Pending, Accepted, Rejected) now updates automatically whenever a review is saved. No manual status management needed.
Why it mattersThe system keeps status in sync across all views automatically, so teams always know which encounters need attention without polling or refreshing.
Improvements
- Smarter encounter list with paginated loading, collapsible adverse event rows, and grade badges (G1 through G5 with severity colors).
- Patient encounters sidebar for quick browsing of all encounters for a specific patient.
- Clinical note panel now shows recent encounters for the same patient (last 30 encounters).
- Design system enforcement across 370+ files: all colors now use semantic tokens for consistent light and dark mode.
- Mobile workflow error recovery: retry failed recordings without losing completed steps.
- Removed 4,000+ lines of deprecated code for a leaner, faster codebase.
Bug fixes
- Fixed deployment build errors.
- Fixed invalid component prop types causing lint warnings.
- Improved audio player accessibility with proper ARIA attributes.
- Fixed navigation path from
/encounter/to/encounters/.
This week's impact
- 18 features and improvements shipped
- 7 bug fixes deployed
- 4,000+ lines of deprecated code removed
- Platform-wide design token enforcement for light and dark mode
- v2026.05
Workflow recovery, clinical data export, and mobile progress tracking
The platform now recovers gracefully from mid-workflow failures, exports clinical session data as a compliance-ready ZIP, and shows real-time step progress on mobile.
gradingmobilerelease12 items shipped15+ screens updated4 bug fixesNew features
Workflow error recovery
When something goes wrong during recording or AI analysis, you no longer get stuck. The app now tracks where you were in the workflow and lets you retry from that exact point without losing your progress.
Why it mattersNetwork hiccups and temporary failures happen, especially in busy clinical environments. Recovering gracefully and continuing from where you left off keeps grading moving without manual re-entry.
Clinical data export
Export your clinical session notes, CTCAE grading history, and clinical decision logs with one click. Everything downloads as a ZIP file ready for your records or compliance documentation.
Why it mattersAudit trails and data portability are critical for clinical trials. A one-click export makes regulatory compliance documentation faster and less error-prone.
Enhanced EHR integrations page
A completely redesigned integrations page with dedicated tabs for EHR systems (Epic, Oracle Health, Athena), communication tools, and productivity apps, now with branded logos so you can see exactly what you are connecting to.
Why it mattersClear visual organization makes it easier to manage clinical tool connections and spot at a glance which systems are active.
ICD-10 code mapping alongside CTCAE terms
When you search CTCAE terms, you now see associated ICD-10 diagnostic codes with confidence scores. This helps with accurate coding and documentation.
Why it mattersAccurate code mapping reduces claim rejections and speeds up billing workflows, so coordinators spend less time correcting downstream documentation.
Mobile workflow progress tracking
The mobile app now shows exactly where you are in the recording and AI analysis workflow. Each step completes visibly in real time.
Why it mattersKnowing where you are in a multi-step workflow reduces uncertainty and makes it clear when you can safely continue with your next patient.
Improvements
- Faster patient search. Now powered by real-time backend search instead of local filtering.
- Consistent mobile styling. All screens use the unified design system for better visual consistency.
- Simplified navigation. Mobile app now uses tab-based navigation: Search, Record, Cases, Settings.
- Better loading states. Skeleton screens show exactly what is loading, reducing perceived wait time.
- Improved dark mode. Fixed contrast issues so text is always readable.
Bug fixes
- Fixed loading indicators not appearing during profile photo upload.
- Fixed adverse event acceptance occasionally not showing confirmation.
- Resolved recording completion sometimes appearing stuck.
- Fixed missing visual feedback on various button interactions.
This week's impact
- 12 features shipped across web and mobile
- 15+ screens migrated to the updated design system
- 200+ lines of styling debt eliminated
- 4 bug fixes for smoother daily workflows
- v0.10
Security, performance, and AI improvements
Big security, speed, and AI fixes.
authreliabilityreleaseSecurity hardeningFaster data loadingImproved AI builderAdded
- New dashboard tools and improved analytics
- More options and better chat for AI builder
Changed
- Optimized response viewer for faster, smoother data loading
- Upgraded dashboards and layouts for a better experience
- Updated many libraries and configs
Fixed
- Major performance fixes across the app
- Issues with AI form builder
- Access and security checks for forms and users
Removed
- Outdated security docs
- Old Black Friday promos
- v2026.04
Unified provider dashboard, multi-event grading, and inter-rater reliability analytics
A new provider dashboard brings session tracking, encounter management, and CTCAE search into one interface. Providers can now grade up to 10 adverse events in a single session, and a dedicated analytics dashboard shows grading consistency metrics your team can share with regulators.
gradingprosafetyrelease5 major features shippedMulti-event grading in one sessionSub-200ms CTCAE searchInter-rater reliability analyticsNew features
Unified provider dashboard
The new provider dashboard brings together real-time session tracking, encounter management, and the full CTCAE search experience in one place. Everything your team needs for a complete patient assessment is available from a single screen.
Why it mattersContext-switching between separate screens for recording, patient management, and grading adds friction to every clinical workflow. The unified dashboard removes that friction so coordinators and PIs can move through an assessment without losing their place.
Multi-event grading in a single session
Grade up to 10 adverse events in one session using a new accordion-style interface. Each event shows AI-suggested grades and confidence scores. You review, adjust if needed, and confirm. Human-in-the-loop review remains the final step on every grade.
Why it mattersClinical coordinators spend significant time grading adverse events one at a time. Multi-event grading lets your team complete a full patient assessment in a single sitting, saving hours per week per coordinator.
Adverse event intelligence with MedDRA classification
The platform now uses MedDRA classification to understand medical terminology hierarchies, group related adverse events automatically, and standardize drug names across your documentation. The grading engine also incorporates drug interaction data and clinical trial evidence when generating suggestions.
Why it mattersMore structured inputs produce more accurate grading suggestions. Grouping related events and standardizing drug names means fewer discrepancies in your safety data across sites and coordinators.
Inter-rater reliability analytics dashboard
A new six-tab analytics dashboard shows grading consistency across your team: agreement rates, classification metrics, confidence calibration, and weekly trends. The data is organized and audit-ready.
Why it mattersFDA requires documentation of inter-rater reliability for clinical trial grading. This dashboard generates that evidence automatically, so your team is not assembling it manually before a submission or audit.
Organization management and team tools
A new settings modal provides team management, bulk invitations, and organization-wide preferences from one screen. A command palette (Cmd+K) is available for power users who prefer keyboard-driven navigation.
Why it mattersGrowing teams need tools that scale. Bulk invitations and centralized permission management make it practical to onboard a full site team without IT involvement.
Improvements
- CTCAE search speed. Results appear in under 200ms with real-time ICD-10 code integration so finding the right term is nearly instant.
- Design consistency. Dark mode support and accessibility improvements applied across all screens.
- Mobile experience. Smoother keyboard handling, faster screen transitions, and feature parity with the desktop experience.
- Audio format support. Additional audio formats are now accepted for voice recordings.
Bug fixes
- Clinical evidence is now editable during grade review (previously read-only)
- Keyboard no longer covers input fields on mobile forms
- Patient creation now properly links to the current visit
- Privacy policy page content is no longer cut off at the top
This week's impact
- 5 major features shipped
- 14 new interface components added to the provider dashboard
- Sub-200ms search response times achieved
These updates are live for all Burna AI users. Questions? Reply to this changelog or email hello@burna.ai
- v0.11
Mobile CTCAE grading and recent cases dashboard
Faster documentation and better data organization.
mobilegradingauthreleaseMobile CTCAE gradingRecent cases dashboard10 updates shippedProduction-ready mobileNew features
Mobile CTCAE grading
Your providers can now grade adverse events directly from the mobile app. Search CTCAE terms, view grade definitions, and submit manual gradings, all from the field or clinic floor.
Why it mattersFaster documentation means more time for patient care.
Recent cases dashboard
See your most recent cases at a glance with statistics showing total AEs, severe AEs, and pending reviews. One-tap navigation gets you into any case instantly.
Why it mattersNo more hunting through lists to find what needs attention.
Personalized experience
The app now greets you by name and remembers your email for faster sign-in. It's the little things that make your day smoother.
Why it mattersLess friction means less frustration during your busy day.
Production-ready mobile
The provider mobile app now connects to our production environment, giving you the same reliability and data you have on web.
Why it mattersReal data, real workflows, ready for real clinical use.
Improvements
Authentication flow overhaul
- Sign-in keeps your email and one-time code entry in context
- Protected routes now show clear login prompts instead of redirecting
- Bottom sheet authentication keeps you in context
Search-focused navigation
- Home tab is now "Search" so you can find CTCAE terms faster
- Branded header with Burna AI logo for clear app identity
- Cleaner, more scannable term lists with better typography
UI/UX polish
- Improved form button styling with proper theming
- Better dark mode colors for easier viewing
- Fixed filter icon visibility issues
- More spacious media selection modal
Bug fixes
- Gender selection now shows relevant options (removed "Unknown")
- Settings screen input fields styled consistently
- Fixed text encoding issues in onboarding screens
- Filter close button is now properly visible
- v2026.03
Mobile CTCAE search and grading, recent cases dashboard, and a production-ready provider app
Providers can now search CTCAE terms, view grade definitions, and submit manual gradings from the mobile app. A new recent cases dashboard shows total and severe adverse events at a glance, and the provider app now connects to the production environment for real clinical workflows.
gradingmobileprorelease10 updates shippedCTCAE search and grading on mobileRecent cases dashboardProvider app in productionNew features
Mobile CTCAE search and grading
Providers can now search CTCAE terms, view grade definitions, and submit manual gradings entirely from the mobile app. The full grading workflow is available from the field or clinic floor, not just at a desktop.
Why it mattersFaster documentation at the point of care means more time for patient care and fewer delays in capturing adverse event data. Grading happens when and where you observe it.
Recent cases dashboard
A new dashboard surfaces your most recent cases with counts for total adverse events, severe adverse events, and items pending review. Tap any case to open it directly.
Why it mattersCoordinators and PIs no longer need to scan through lists to find what needs attention. The dashboard puts the most urgent cases front and center so triage takes seconds, not minutes.
Production-ready provider mobile app
The provider mobile app now connects to the production environment, giving your team the same data and reliability on mobile that they have on web.
Why it mattersReal data and real workflows are now available wherever your team is. Mobile is no longer a secondary experience.
Personalized sign-in experience
The app now greets you by name and remembers your email for faster sign-in on return visits.
Improvements
Authentication flow
- Sign-in keeps your email and one-time code entry in context
- Protected areas of the app now show a clear sign-in prompt instead of silently redirecting
- Bottom-sheet authentication keeps you in context without losing your place
Navigation
- The home tab is now labeled "Search" to match how most providers actually use it
- The Burna AI logo is prominent in the header for clear app identity
- Term lists are cleaner and more scannable with improved typography
Interface polish
- Improved form button styling with consistent theming
- Better dark mode colors for easier viewing in low-light settings
- Filter icon visibility issues resolved
- Media selection modal is more spacious and easier to use
Bug fixes
- Gender selection now shows relevant options only
- Settings screen input fields styled consistently across all forms
- Fixed text encoding issues on onboarding screens
- Filter close button is now properly visible
This week's impact
- 10 updates shipped
- 18 features added
- 7 bugs resolved
Questions about these updates? Reply to this changelog or email hello@burna.ai
- v2026.02
Mobile grading, AI-generated clinical notes, and real-time workflow results
Providers can now manually grade adverse events from a phone or tablet, audio recordings automatically produce structured clinical notes with diagnosis and billing codes, and workflow results update in real time without blocking the interface.
gradingmobileprorelease3 major features shippedMobile CTCAE grading liveAI clinical note generationReal-time workflow resultsNew features
Manual CTCAE grading on mobile
Providers can now manually add and grade adverse events directly from their phone or tablet. CTCAE grades, notes, and documentation sync instantly to the rest of your team whether you are at the bedside, in a hallway, or in a meeting.
Why it mattersAdverse events do not wait for desk time. Grading the moment you observe an event means more accurate documentation and fewer end-of-day catch-up sessions.
AI-generated clinical notes from audio
Upload a visit recording and the platform automatically produces a structured clinical note with ICD-10 diagnosis codes and CPT billing codes extracted and ready for review. The note is generated from the audio, not entered by hand.
Why it mattersDocumentation time that used to take hours compresses to minutes. Clinicians focus on the patient during the visit and review a structured note afterward rather than writing it from scratch.
Workflow history tracking
Every AI-assisted grading session, transcription, and analysis is now properly linked and auditable in a dedicated history view. Workflow associations are stored in a flexible architecture that keeps your visit records clean and complete.
Why it mattersA clear audit trail for every automated action is a regulatory requirement. Workflow history gives your team and your auditors a single place to see what happened, when it happened, and which visit it belongs to.
Improvements
Real-time workflow results
Workflow results now stream in as they are ready. You can continue working in the platform while results arrive, with no page refresh needed.
Mobile performance
- Optimized data streaming for lower latency on mobile networks
- Improved safe area handling so screen content no longer overlaps device notches on newer iPhones and Android devices
- More intuitive and consistent date picker across iOS and Android
- App size reduced: storage utilities consolidated for a 2.1 MB smaller install
Patient management
Patient profiles are now properly scoped to your organization with enhanced multi-tenant isolation, so each team's data remains separate and secure.
Bug fixes
- Theme colors are now consistent across all screens
- Audio recordings over 10 minutes now transcribe reliably
- Workflow status updates appear immediately without requiring a page refresh
This week's impact
- 3 major features shipped
- Mobile grading available to all provider app users
- AI clinical notes in early access for audio upload workflows
Have thoughts on these updates? Reply to this changelog, email us at feedback@burna.ai, or chat with us in the app.
- v2026.01
Model transparency for AI grades, automatic AE categorization, and email authentication
Every adverse event grade now shows which AI model contributed to the assessment, system organ class categorization happens automatically, and teams can now sign in with email and password for institutions that require it.
gradingcomplianceauthrelease5 features shippedAI model tracking on every gradeAutomatic organ class categorizationEmail and password sign-inNew features
AI model tracking on every grade
Every adverse event grade now displays which AI model generated the assessment. The model identifier is part of the grade record and carries forward into your audit trail and any downstream reports.
Why it mattersRegulatory bodies require clear documentation of AI involvement in clinical assessments. With model tracking, your team can demonstrate exactly which system contributed to each grading decision, supporting 21 CFR Part 312 compliance requirements and giving auditors a complete picture.
Automatic adverse event categorization by system organ class
Adverse events are now automatically sorted into their system organ class: hematologic, neurologic, dermatologic, gastrointestinal, cardiovascular, pulmonary, renal, hepatic, metabolic, infectious, and other categories. No manual categorization is needed during patient visits.
Why it mattersConsistent categorization across your team makes it far easier to spot safety patterns, such as all GI toxicities associated with a specific drug regimen. It also saves coordinators and PIs time during assessments by removing a step that was previously done by hand.
Simplified authentication with email and password
You can now sign up and sign in using a traditional email address and password, in addition to social authentication options. Both paths are available at sign-in.
Why it mattersHealthcare organizations often have specific authentication requirements for compliance. Password-based authentication ensures every team member can access the platform regardless of their institution's IT policies around social login.
Automated waitlist processing
When prospects submit the waitlist form, the system now researches their background, generates a personalized draft response, syncs the contact to the CRM, and sends an internal notification. Your team gets instant context about each prospect's clinical trials experience.
Why it mattersResponse time to new interest compresses significantly. Your team spends time on the conversation, not on manual research and data entry before it can start.
Improvements
Mobile app
- Fixed authentication state synchronization issues that occasionally caused incorrect redirects
- Improved safe area handling on modern iOS devices
- Enhanced app responsiveness with better state management
Billing
- Streamlined subscription management with improved customer account tracking
- Simplified billing workflows for enterprise customers
Performance
- Reduced unnecessary data fetching by consolidating user profile lookups
- Faster page loads when accessing user information
This week's impact
- 5 features shipped
- 4 customer-facing improvements deployed
- Multiple authentication and mobile experience enhancements
- Enhanced regulatory compliance capabilities
Have feedback on these updates? Reply to our newsletter or email us at hello@burna.ai
- v2025.52
Automated CTCAE grading, e-signatures, and enterprise audit infrastructure
Burna AI now grades adverse events automatically with a single click, attaches cryptographic signatures to every AI-generated grade for 21 CFR Part 11 compliance, and moves audit records to a six-year retention store that meets FDA and sponsor audit requirements.
gradingcompliancesafetyrelease3 major features shippedAutomated CTCAE grading live21 CFR Part 11 e-signatures6-year audit retentionNew features
Automated CTCAE grading
Clinicians can now trigger AI analysis of adverse events with a single click. The system evaluates patient data, matches events to CTCAE v6 criteria, and returns grades with confidence scores and detailed rationales. Human-in-the-loop review remains the final step; the AI suggests, clinicians decide.
Why it mattersGrading time drops from 15 to 20 minutes per adverse event to seconds. Coordinators and PIs spend their time on clinical judgment, not manual look-ups. The system is also designed to support blinded grading studies that compare AI performance against expert clinician review.
Grading performance analytics
A new analytics dashboard gives your team real-time visibility into grading performance: agreement rates between AI suggestions and human reviewer decisions, confidence calibration trends, and override rates across your study portfolio.
Why it mattersQuality assurance for clinical trial operations now runs continuously. Your team can see where AI suggestions are most reliable and where additional human review is warranted, without waiting for a manual audit cycle.
Enterprise audit infrastructure with six-year retention
Every action in the platform is now logged with HIPAA-compliant storage and real-time user enrichment for complete traceability. Audit records are retained for six years, meeting FDA submission requirements and sponsor audit standards.
Why it mattersRegulatory bodies expect a complete, tamper-evident audit trail for all clinical data operations. This infrastructure makes FDA submissions and sponsor audits significantly less painful by providing records that are already structured for review.
Security and compliance
21 CFR Part 11 e-signatures on AI grades
Every AI-generated grade now carries a cryptographic signature using SHA-256 hashing. Each signature is tamper-proof and includes complete provenance tracking: which model contributed, which clinician reviewed, and when.
Why it matters21 CFR Part 11 requires electronic signatures on any record that would otherwise require a handwritten signature. Attaching signatures at the grade level means your audit trail is compliant by construction, not by policy.
HIPAA compliance hardening
Zero protected health information in system logs, automated PHI exposure prevention, and IP spoofing protection are now enforced at the platform level.
Improvements
- Enhanced IP address validation prevents security vulnerabilities in audit records
- Standardized error handling gives clearer feedback when operations fail
- Improved confidence score handling eliminates data quality gaps
- Better authentication reliability with updated security packages
Bug fixes
- Fixed an authorization issue that could cause null return errors
- Eliminated PHI logging violations to maintain HIPAA compliance
- Corrected audit record type mismatches for reliable data access
This week's impact
- 3 major features shipped
- 150+ tests passing, including 18 new comprehensive tests
- 60+ legal documents prepared for pilot partnerships
- v0.8
Changelog page and UI improvements
New changelog system, header redesign, and hero updates.
marketingreleaseNew changelog pageRedesigned headerMobile-responsive layoutAdded
- Changelog page with markdown-based changelog system
- Individual changelog entries with frontmatter support
- Sticky date layout for changelog entries
- Responsive changelog layout for mobile and desktop
Changed
- Redesigned header section for better UX and reduced layout shift
- Updated Vercel brand icon in hero section
- Improved header navigation structure and styling
Fixed
- Reduced layout shift in header component
- Enhanced header accessibility and user experience
- v1.2
Enhanced AI agent interface and reasoning capabilities
Major improvements to AI agent performance, conversation interface, and reasoning abilities.
gradingreliabilityrelease60% faster responsesStronger reasoningBetter context awarenessImprovements
- Enhanced reasoning with improved logical thinking and problem-solving
- Context awareness with better understanding of conversation history
- Multi-modal support to process text, images, and code simultaneously
- Adaptive responses tailored to communication style based on user preferences
- Real-time learning with dynamic adaptation during conversations
- v0.9
Interface redesign and accessibility improvements
A cleaner, more intuitive interface with stronger keyboard navigation, screen reader support, and responsive layouts across device sizes.
mobilereleaseRedesigned navigationStronger accessibilityResponsive on every deviceImprovements
- Redesigned interface with cleaner, more intuitive navigation
- Enhanced accessibility with improved keyboard navigation and screen reader support
- Streamlined workflows for faster task completion
- Responsive design optimized for all device sizes