The situation
Prospects hesitate because they cannot picture the result, and clinics spend hours answering the same first questions. An AI preview helps only if it never reaches a patient unchecked.
AI agent · Health
See the result before the procedure, approved by the clinic
A hair-transplant clinic wanted prospects to picture their result before booking. We built the clinic's website, its back end, and an AI estimator that turns a patient's photos and answers into a projected-result card, reviewed by the clinic before it is sent.
Illustration
AI agent · Health
- Photos and answers
- Clinic accepts
- Project the result
- Assess and recommend
- Clinic sends
What changed
The patient submits photos and a few answers. After the clinic accepts the application, an image model produces a projected-result card and a text model writes a short assessment. The technique and package follow from a fixed rule, and nothing is sent until the clinic has reviewed it.
How it works
5 steps, end to end
- 01
Photos and answers
The patient uploads photos and describes their hair-loss stage and zones.
- 02
Clinic accepts
The application waits for the clinic's approval before any AI runs.
- 03
Project the result
An image model generates a projected-result card.
- 04
Assess and recommend
A short written assessment; the technique and package follow a fixed rule.
- 05
Clinic sends
The clinic reviews the card, then sends the patient their link.
Where the AI stops
The model writes. The code decides.
The model
- Generates the projected-result image
- Writes a short assessment
The code
- Runs nothing before the clinic accepts the application
- Derives the technique and package from a fixed rule
- Records the real cost of every generation
- Falls back to a set text if the model fails
By the numbers
Counts read in the code we shipped. No invented results.
- 2
- human approvals before a patient sees anything
- 5
- states every request moves through
- 1
- fixed rule for the recommendation
Built with
- Image and text models
- Job queue and scheduler
- Postgres and storage
- Bot protection
- Email workflow
More use cases
All use cases
- AI agentsA legal assistant that cites the law, or says it doesn't knowA legal assistant that only answers from the law it retrieved, and checks every article it cites.SAAFKA
- Voice AIA sales trainer that picks up the phoneTrainees phone a realistic AI property owner, then get a scored debrief with one action to work on.ConciergElite
- AI agentsA coach that reads the playbook so advisors don't have toA coaching agent that retrieves only the right sections of a large knowledge base, and keeps internal names out of its answers.SAAFKA
Contact
let's talk about your project
- Where we areMorocco, working worldwide
- Writehello@kaishift.com
- LanguagesFrench, Arabic, English
Write to usWe answer the same working day.