AI engineer
- Format
- Remote
- Employment
- Full-time
- Salary range
- ₽250k–400k
You will own rollouts end to end: from taking the process apart and cleaning up the knowledge base to quality metrics after launch. The person who designs it here is the person who fixes it.
Tasks
- Take the client's process apart and work out where an agent pays for itself
- Build the knowledge base and document search with the source cited
- Connect tools: CRM, calendar, internal APIs
- Build evaluation sets for quality and watch how they move
- Test our own agents for prompt injection before handing them to the client
Requirements
- Python at the level of production code, with tests and types
- Experience putting at least one LLM solution into operation
- An understanding of RAG: chunking, re-ranking, quality evaluation
- Willingness to tell a client that their problem is not solved by a model
Nice to have
- Experience with Langfuse or another observability tool for LLMs
- Familiarity with attacks on LLMs and the ways of defending against them
- Experience deploying open models inside a closed perimeter
Conditions
- Remote from anywhere, with 4 hours of overlap with Moscow time (MSK)
- AI tools and APIs paid for in full by us
- A training budget of ₽100,000 a year
- A share of the projects you ran from brief to handover
- Flexible hours, and no time tracker
ApplicationTell us about yourself