Skip to content
← All work

Health AI · Clinical training

Clinical training simulator

A clinical training MVP in development: students work through cases, guided by a state machine and evaluated with AI feedback and scoring. Several modules are functional; the platform is not fully delivered.

MVP in developmentSeveral modules are functional. The platform is not presented as fully delivered.

clinical-sim / session loop
Illustration
Illustration of a training session loop: student action, state transition, clinical evaluation and AI feedback, producing a score.

The problem

Clinical training needs structure: cases must move through controlled steps and feedback must be consistent. An open-ended chat is not enough.

Context

A training MVP in development, where a student works a clinical case step by step instead of talking to an unconstrained chatbot.

My role

I designed and built the session loop now in the MVP: backend orchestration, the state machine that governs case steps, the clinical evaluation of student actions, and the AI feedback and scoring that close the loop.

What I built

Built features

  1. 01Case presentation and student actions
  2. 02Controlled step transitions
  3. 03Evaluation of actions against case logic
  4. 04AI-generated feedback
  5. 05Session scoring

Design decisions

  • Drive the case with a state machine so only allowed transitions can happen.
  • Keep orchestration, evaluation and feedback as separate backend responsibilities.
  • Score the session from evaluated actions, not from an open-ended conversation.
  1. 01A backend orchestrator coordinates each training session.
  2. 02A state machine governs case steps and allowed transitions.
  3. 03A clinical engine evaluates the student's actions.
  4. 04AI feedback and scoring close the loop.

Outcome

Several modules of this clinical training MVP are functional. The platform is not presented as fully delivered.

Architecture

How Clinical training simulator fits together.

Each step, in order. Select a node to see what it is responsible for.

  1. A student working through a clinical case.

  2. The interface where the case is presented and actions are taken.

  3. Coordinates each step of a session across the backend components.

  4. Controls which step of the case is active and which transitions are allowed.

  5. Evaluates the student's actions against the case logic.

  6. Generates feedback on the student's decisions.

    • LLM

  7. Produces the session score.

Stack

  • Orchestration
  • State machine
  • Clinical engine
  • LLMs

Next project

AI content generation platform

An internal product in development: AI features that turn a brief into generated marketing content, grounded in retrieved context.