I take AI systems from prototype to production
I am Léonard Tschora. I build and ship AI systems: RAG and agent architectures, LLM pipelines, and forecasting models that a decision actually depends on. I trained on European day-ahead power markets, where a bad model loses money by lunchtime. I work with companies as an independent consultant, remote from Paris, France.
What I do
GenAI to production
RAG and agent architecture, evaluation harnesses, LLM pipelines, and the cost and latency work that turns a prototype into something which survives real traffic.
Forecasting and time-series
Demand, price, load, capacity. Feature and validation design, and putting the model where a decision depends on the number it produces.
Private and on-premise LLM
Self-hosted model deployment and private retrieval over internal data, for organisations whose data cannot leave their perimeter.
Three applications you can check
- Yes we kanjiA web app for memorising Japanese kanji.
- ChoCacaoA web app that finds the hottest and coolest communes in metropolitan France for any hour in the week ahead.
- gengiscanA Discord bot that turns a phone photo of a handwritten page into structured Markdown, committed straight into the user's GitHub notes repository.
Start with a conversation
Thirty minutes, no deck. Tell me what you are building and what is stopping it from shipping, and I will tell you whether I am the right person and roughly what it would take. If I am not, I will say so.
or email leonard.tschora@protonmail.com