AI engineer

Who you will work with

Our AI agents read documents, find an internal procedure, draft an answer and wait for a person's approval before they act. You will build them with us, from the choice of model to their launch at the client, and you will know how to measure when they get it wrong, because a company only keeps an agent that is as reliable on the hundredth file as on the first.

What you will do

• Design agents and document search pipelines (RAG) on our clients' data.
• Pick the right model for each need from those on the market, weighing answer quality, speed and running cost.
• Build the test sets and evaluations that show what an agent gets right and what it misses.
• Connect agents to the client's tools with only the rights their task needs, and plan human approval for sensitive actions.
• Monitor agents once they are live, their errors and their response times, and improve them.
• Explain your choices to the client's teams in plain words.

Your profile

• You have at least three years of software development behind you, part of it on applications built on language models and used in production.
• You are strong in Python and comfortable in TypeScript.
• You know vector databases, embeddings and document chunking, and you can tell why a search missed the right page.
• You have set up model evaluations before and you read their results with care.
• You know the Model Context Protocol or tool calling, and you follow each new generation of models closely.
• You work in French and English.

Tools for the role

• Claude (Anthropic)
• OpenAI models
• Qdrant
• MCP
• Python
• TypeScript
• n8n
• PostgreSQL
• Docker

Nice to have

• Experience with document reading (OCR, vision).
• Arabic or Darija, useful for our clients' cases in Morocco.
• A side project or a publication about agents.

How to apply

Write to careers@clarodigi.com with “AI engineer” in the subject line, your CV and a few samples of your work. If your profile is a match, we will invite you to an interview.