← Home
Business system · AI · healthcare
AI-powered medical CRM for a premium functional-medicine clinic
A private system for a nutrition and functional-medicine clinic. Doctors keep a complete patient record, and AI helps analyse it. No AI output enters the chart without a doctor's approval. Patient data stays on the clinic's own server, not in the cloud.
Features
- Patient record: history, ICD-11 diagnoses with a full amendment trail, medications and supplements with doses and units, lab results with trend charts, genetics, visits and documents.
- Lab PDF parsing: AI drafts the values, and a doctor checks them against the original and approves them.
- Knowledge-grounded AI advice that draws on the clinic's own knowledge base and shows its sources.
- AI research swarm: a team of local AI agents, each tuned for its own task (search, analysis, critique, reporting), runs research on given topics under the supervision of top models Claude and ChatGPT — searching PubMed, Europe PMC and ClinicalTrials.gov and writing a report with real citations.
- Your own local knowledge base for training the AI agents: protocols, articles, approved studies and the clinic’s experience stay on its server and become the agents’ memory — no data leaves the clinic.
- Switchable AI engines: OpenAI GPT, Anthropic Claude or a local MedGemma model, with API keys stored encrypted.
- One-click installer for a Mac server, encrypted backups and an automated self-test.
- Three UI languages: Ukrainian, Russian and English.
Stack
- Next.js 14
- React
- TypeScript
- PostgreSQL / Supabase (RLS, Vault, pgvector)
- Docker
- OpenAI API
- Anthropic API
- Ollama
- RAG
By the numbers
78 database migrations, 520 automated tests, access checks for every user role, and an audit trail for every action on clinical data.
Screenshots
Screenshots were taken on a demo setup: all patients, doctors and lab results are fictional. Built for a clinic; the client’s name is not disclosed.






Need a similar system?
We will deploy AI agents in your company — locally or in the cloud.