Studio Beyowi
Python Software Developer — Apprenticeship
October 2023 — September 2026
- Healthcare — Developed most of the Thésorimed backend, including medicine-document processing.
- Real estate — Property valuations with the CTO using cleaned DVF transactions in PostgreSQL.
- Data — Medicine-document, price and reimbursement ETL; SQL migrations with AWS DMS.
- RAG & OCR — Queries over medicine leaflets/SmPCs and user-provided real-estate information.
- Delivery — AWS, DigitalOcean, Docker, CI/CD, n8n; development, staging and production.
Examples of contributions
Primotheus: value properties using comparable transactions
I contributed alongside the CTO to the FastAPI backend of a real-estate platform. DVF transaction data is cleaned and stored in PostgreSQL. Valuation retrieves comparable sales by location, property type and floor area, filters outliers and calculates an estimated value. The backend also supports refreshing valuations after property changes or through a batch process.
Thésorimed: automate medicine-data preparation
I developed most of the Flask backend for the internal Thésorimed platform. Processing retrieves and updates medicine leaflets and summaries of product characteristics (SmPCs), extracts their sections and prepares data for SQL databases. Pandas pipelines also process medicine prices and reimbursement data; documents and extraction outputs are stored in AWS S3.
RAG: query relevant documents in context
In healthcare, I contributed to RAG for questions over medicine leaflets and SmPCs. For Primotheus, RAG queries real-estate information supplied by the user: their properties and documents, including content extracted through OCR. These contributions combine vector search and OpenAI services; the real-estate use case aims to work with each user’s own supplied data.
Primotheus: connect the backend to n8n workflows
The Primotheus backend integrates with n8n workflows to process events and synchronize information with business tools. This provides context for my Python, FastAPI, REST API and data-processing work, alongside pytest tests and delivery through Docker and CI/CD.