Launching soon — in development and not yet validated. Research pilots are being scoped with a small number of health professions education programs.
Document feedback in the moment
The AI Feedback Scribe is designed to help learners and instructors capture short feedback conversations from clinical training and turn them into structured, human-reviewed assessment evidence.
Post-encounter debrief
4:12 · Clinical communication
Instructor
“Your history was focused and you explained the plan clearly. Next time, make your diagnostic reasoning more explicit.”
Learner
“I wasn’t sure how much of the differential to include.”
Instructor
“Let’s capture that as a next step.”
AI draft
Needs reviewMini-CEX · Clinical communication
- Strengths
- Focused history; clear explanation of the plan
- Next steps
- Make diagnostic reasoning explicit
The documentation gap
Most feedback is spoken, then lost
Clinical training produces a steady stream of useful feedback. Much of it happens during or immediately after clinical work — and is never written down.
-
Faculty reconstruct from memory
By the end of a rotation, an instructor may be asked to recall details from conversations that took place days or weeks earlier.
-
Learners receive a brief summary
A rating and a sentence or two may be all that remains of a substantive discussion about performance.
-
Assessment teams lack timely evidence
Competency committees need specific, dated examples of learner development — not a retrospective impression.
The AI Feedback Scribe is designed to close that documentation gap, without replacing the conversation or the educator’s judgment.
How it works
From a short conversation to usable evidence
The intended workflow, built around the assessments your program already uses.
- 01
Record a short debrief
A learner or instructor captures the feedback conversation on mobile or web while the details are fresh.
- 02
Transcribe by speaker
The conversation becomes a speaker-labelled transcript.
- 03
Draft and map
The scribe drafts the evaluation and suggests relevant competencies, EPAs or assessment fields.
- 04
Review and approve
Learner and instructor correct and approve the content before submission.
- 05
Build the longitudinal view
Approved assessments contribute to dashboards and competency review.
AI assists with documentation. It does not replace educator judgment or the required human review.
Capabilities
Built for the feedback that already happens
-
Ambient feedback capture
Record short learner-instructor debriefs while the details are still fresh.
-
Automatic transcription
Convert the conversation into a speaker-labelled transcript.
-
AI-drafted evaluations
Use the transcript to populate existing assessment forms rather than requiring faculty to document the conversation again.
-
Competency and EPA identification
Suggest relevant competencies, EPAs and assessment categories based on the discussion.
-
Flexible assessment workflows
Support workplace-based assessments such as Mini-CEX, DOPS, ITER, EPA forms, field notes and program-specific evaluations.
-
Human verification
Allow learners and instructors to review, edit and approve the generated content.
-
Continued dialogue
Support follow-up discussion when additional context or clarification is needed.
-
Longitudinal competency insights
Help authorized users explore trends, strengths and areas for improvement across multiple assessments.
-
Natural-language questions
Let authorized users ask questions in plain language. For example, a competency committee might ask: “How have this learner’s communication scores changed over the past six months?”
Feedback quality rating
Help faculty learn to provide better feedback by comparing feedback transcripts to feedback models like ADAPT or CPR2.
Research partnership
Help us measure whether better capture leads to better feedback
Resilience Software is seeking a few programs to conduct collaborative research pilots of the AI Feedback Scribe.
The research question
Does making feedback easier to capture lead to more feedback, sooner — and to better evidence for competency decisions?
Pilots may be configured around your existing Mini-CEX, DOPS, EPA, field-note or other workplace-based assessment workflows.
We are not claiming that the AI Feedback Scribe improves educational outcomes. The purpose of these pilots is to find out.
The objective is to evaluate whether the AI Feedback Scribe improves:
- The quantity and completion rate of workplace-based assessments
- The timeliness of documented feedback
- The specificity and actionability of feedback
- The amount of narrative evidence available for competency decisions
- Learner and faculty satisfaction
- Perceived documentation burden
The institution leads
- Leads or co-develops the research protocol
- Obtains its own ethics, privacy, consent and administrative approvals
- Defines the evaluation questions that matter to its program
Resilience supports
- Product configuration and onboarding
- Assessment-form setup
- Appropriate technical and security documentation
Governance
Designed for Responsible Institutional Evaluation
Research pilots are intended to fit within your institution’s existing oversight, not work around it.
-
Program data ownership
The institution retains ownership of its program data.
-
Clear roles
Resilience Software acts as the technology provider and data processor.
-
Human review
Human review remains part of the assessment workflow.
-
Institutional requirements
Pilot implementation is subject to institutional privacy, security, ethics and consent requirements.
-
Review documentation
Security, privacy, AI-use and third-party provider documentation can be supplied during institutional review.
Launching soon
Tell us about your program and the feedback workflow you want to study. We will scope a pilot around your existing assessments.