ML Tech Lead (LLM Agents)

tldr: ML Tech Lead (LLM Agents), 700k gross, remote worldwide

Context Luchi is an online insurance company (Voluntary Medical Insurance, >1 million insured people) with its own telemedicine clinic. We are building AI medicine: an agent-based system that handles routine tasks for our doctors and operators — collecting anamnesis and context for appointments, summarizing medical records, drafting conclusions, and checking prescriptions. The team (3 ML, 2 BE, SA + QA) is already assembled — we need a Tech Lead for the team: architecture, key technical solutions, code review, and bringing the system to production. People management (hiring, 1:1s, performance reviews) is not part of the role but can be included if desired.

Responsibilities Next 3 months: Bring the anamnesis collection pilot to production in the clinic — agent architecture (LangGraph), orchestration and context management, evaluation loop with doctor annotation, quality and cost metrics per request. 6–12 months: Expand the system to medical record summarization and draft conclusions, set up a fine-tuning loop on clinic data (SFT/LoRA), and scale the solution to the entire clinic while controlling economics.

What we need — 4+ years in ML, with LLM solutions deployed to production under real load; — RAG, agent architectures, MCP; fine-tuning (LoRA/QLoRA/SFT); — Python, FastAPI, Docker/k8s, Kafka/RabbitMQ; Plus: MedTech, NLP on medical texts, vLLM/SGLang in production.

Stack: Python 3.13, FastAPI, PostgreSQL, LangGraph, MLflow, Kafka, k8s.

Format: full-time, employment contract / civil law contract, Russian legal entity. Remote from anywhere in the world or hybrid (office — Moscow, m. Savyolovskaya). Range: 600–700 thousand rubles gross + Luchi VMI for self and children, 28 days of vacation.

Contact: Artem Derut, @artemderut