Athletes have replays.

Pilots have recorders.

Engineers have logs.

Doctors… what do doctors have?

Now doctors have

Review today. Save lives tomorrow.

Scroll

The Problem

Nobody sees the mistakes.
Not even the student making them.

  • 2,000+ new medical students every year1
  • ~20 hospitals where clinical training happens2
  • 400 students training in one hospital at once3
  • A few rushed minutes of feedback per student, on a lucky day

Consultants rightly put patients first. A ward round is not a classroom, so guidance is scarce, for good reasons.

Students see patient after patient, and the details blur. A step skipped on Tuesday is skipped again on Friday, because nobody saw it happen, including the student.

How might we improve the training process of a Sri Lankan medical undergraduate?

Problem Statement
  1. 1 2,049 entered state medical schools in 2024; UGC intake for 2022/23 was 2,085.
  2. 2 Counted as hospitals where clinical training takes place, which is broader than the officially designated “Teaching Hospital” list.
  3. 3 Roughly 100 students per hospital per batch, with four batches training at the same time.

The Solution

Give the student the replay.

Phase one Available in the concept as designed

  1. 01

    Record.

    Smart glasses capture the patient interaction exactly as the student sees it, with the patient’s consent, every time.

  2. 02

    Review.

    Timestamped, searchable footage. Rewatch the examination, spot what you missed, time yourself.

  3. 03

    Share.

    Share a recording with the batch, and one student’s patient becomes two hundred students’ lesson.

Phase two The AI layer: where we are heading next

Real-time nudges

  1. Audio. Miss a step and a gentle sound says something’s off, the way a supervisor clears their throat. It doesn’t say what. The student thinks first.

  2. Visual. Still unsure? A glance to the side shows the exact missed step.

Daily feedback

Each evening MedFlix highlights the moments that went wrong, shows progress, and says what to work on tomorrow.

A guide beside you during the examination.
A mentor reviewing your footage after.

The Product

What it actually looks like.

The idea, in 20 seconds

The whole loop in one take the nudge in the moment, then the replay after.

The mobile app

Review recorded interactions wherever you are, revisit feedback, and keep track of what to improve before the next patient.

The MedFlix dashboard: a greeting reading Good evening, Amara, a featured respiratory examination recording, interaction analytics showing duration, feedback score and checklist completion, and Netflix-style thumbnail rows titled My interactions and Shared with me.

The dashboard

Every interaction you have recorded, and every one your batch has shared, laid out as rows you can scan in seconds. Duration, feedback score and checklist completion, tracked over the month, not guessed at.

A medical student wearing smart glasses examines a patient’s abdomen while sound waves ripple from the glasses’ speaker: the audio nudge that something is off.
First, the audio nudge
The student glances aside and a magnified view of the lens shows the exact step, a missed step alert reading Check for scars.
Then, the visual cue

Through the glasses

First the sound: something’s off. If the student is still unsure, a glance aside shows the exact step, and their eyes go back to the patient.

Two devices, one student

A first-year needs a supervisor’s voice in the moment; a final-year only needs the review at the end of the day. So the hardware follows the student.

Meta Ray-Ban Display smart glasses, front three-quarter view, with a camera in the corner of each frame.
Starting out Meta Ray-Ban Display

Camera plus an in-lens display. The audio alert and the glance-up cue in the moment, the review at night.

AIsee, a lightweight head-worn band with a forward-facing camera on one arm and open-ear speakers, with no lenses and no display.
With experience AIsee

Camera and open-ear audio, no display. By now the daily feedback is all they need.

Assisting goes down as experience goes up. That was on the flipchart before it was on a roadmap.

Who it’s for

Sri Lankan medical undergraduates.

Starting with early adopters at the University of Moratuwa, whose young medical faculty has the fewest built-up resources, and the most to gain.

How it rolls out

Trials first. Then the council.

  1. 1 Trials with selected hospitals and universities.
  2. 2 The Sri Lanka Medical Council, so MedFlix becomes part and parcel of medical training.

Licensed to institutions, kept affordable for students.

Data protection

Consent first.

Recording a patient is a serious act. MedFlix is being designed around Sri Lanka’s Personal Data Protection Act, and around one rule before any other:

  • Consent

    Nothing will be recorded without the patient’s explicit consent for a stated purpose. Where the patient is a minor, a guardian’s.

  • SL PDPA

    Health data is a special category under the Act. Its requirements on security, retention and patients’ rights will shape how MedFlix is built.

How we got here

Our Journey

The half above is the product. This half is the rooms, flipcharts and arguments that got us there. Everything below happened.

  1. 01

    We listened.

    We interviewed around ten medical students across three universities: Moratuwa, Colombo and Sri Jayawardenapura. We expected to hear about equipment. We heard about being alone.

    ~10students interviewed
    3universities

    “If I can review my own work, that is a game changer.”

  2. 02

    We framed it.

    Medical education runs lecture, then training, then evaluation. Only the middle is unsupervised. Training, where students are most alone, is where we aimed.

    Problem Statement V2

    Lecture
    Training
    Evaluation

    Assisting + Evaluating

    • Guide the student during training with feedback.
    • Record & review the patient interaction, and share it among students, with patient consent.
    • Indicate time elapsed per examination.
    • Timestamp necessary events for further analysis.
    • Assisting ↓ as experience ↑.
    The original flipchart photographed in the room: Problem Statement V2, the Lecture to Training to Evaluation diagram, and the handwritten bullet list beneath it.
    The original, in the room
  3. 03

    We mapped the student.

    A Value Proposition Canvas, built from the interviews. The left side holds only what the student’s pains demanded, nothing we simply wanted to build.

    The original hand-drawn Value Proposition Canvas on paper, with the square value map on the left and the circular customer profile on the right, annotated in red and blue.
    The original, in the room

    Value Proposition Canvas

    What we build

    Features

    • Recording patient interactions
    • Show timer
    • Supervisor-like pointers

    Gain creators

    • Effective self-learning
    • Recorded interactions shared among students
    • Goal-based, timely reports

    Pain solvers

    • Record the interaction and review it later
    • Can show time without the patient knowing
    • Pointers reduce unintentional overlooks

    Who we build for

    Gains

    • Review patient interactions
    • Share interactions with other students
    • Feedback on my work and progress

    Jobs

    • Take patient history
    • Examinations
    • Study human anatomy

    Pains

    • Less guidance
    • Forgetting patient interactions
    • Can’t time without patient knowledge
    • Unintentional overlooks
  4. 04

    We stress-tested the idea.

    A Lean Canvas, to find where the idea broke. It mostly held. The existing alternative, reviewing notes with doctors, fails only because doctors have no time.

    The original hand-drawn Lean Canvas on paper, nine boxes filled in with blue pen, numbered one to nine.
    The original, in the room

    Lean Canvas

    2 Problem

    • Less guidance during training, and unintentional overlooks
    • Forgetting patient interactions
    • Can’t time without the patient knowing

    Existing alternatives. Doctors in teaching hospitals are time-constrained, so students review their notes with doctors when they can.

    4 Solution

    • Nudge users like a supervisor in real time, and provide daily feedback
    • Record patient interactions for later review
    • Show elapsed time using audio or visual cues

    3 Unique value proposition

    Review the first-person perspective of patient interactions, built for medical students.

    9 Unfair advantage

    A complementary team with first-hand exposure to the target audience.

    1 Customer segments

    • Sri Lanka Medical Council
    • Sri Lankan medical undergraduates

    Early adopters. University of Moratuwa medical students. The faculty started recently, so there are fewer built-up resources.

    8 Key metrics

    • Key action: recording patient interactions
    • Success metric: university adaptability

    5 Channels

    • SLMC
    • Medical Student Union
    • Word of mouth

    7 Cost structure

    • Wearable
    • Hosting (analyser)
    • Storage
    • LLM usage

    6 Revenue streams

    • Subscription for students
    • Licensing fee for SLMC and universities

    The original canvas also carried the day’s working figures, 2,095 students and 22 hospitals. The headline numbers above use the more conservative wording.

  5. 05

    We chose the hardware.

    Glasses, because the camera has to see what the student sees. But one flipchart line kept coming back: assisting goes down as experience goes up. So one device cannot be right for all five years.

    Early years Meta Ray-Ban Display

    Glasses with an in-lens display. The student is still learning the steps, so they get prompted while it matters, and reviewed afterwards.

    Later years AIsee

    No glasses. The student no longer needs prompting mid-examination, so the device drops back to capture and the daily feedback they still use.

    The product does not lose features. The student outgrows them. Recorded on the flipchart as “Meta Ray-Ban Display → AIsee V2”.

  6. 06

    We said it out loud.

    The pitch, as delivered on stage. One camera in the room, one take.

Team

Built by a team with first-hand exposure to the problem.

Mentored by

Talk to us

Want to bring MedFlix to your hospital or faculty?

Curious about the project, the research behind it, or where it goes next? We’d love to talk.

saai@designthinking.lk