GenAI Engineer

As a GenAI Engineer at TMLC, you'll build the LLM-powered products our clients put in front of real users — knowledge assistants, document intelligence, copilots and multi-step agents. The hard part isn't calling an API; it's making these systems accurate, grounded and reliable at enterprise scale. That's the work.

You'll design retrieval pipelines, engineer prompts and agent flows, and build the evaluation harnesses and guardrails that keep them honest. You'll care as much about hallucination rates and latency as about the demo looking good.

What you'll do

• Build RAG systems end to end — chunking, embeddings, vector search, reranking and grounded generation with citations.
• Design agentic workflows — tool use, planning and multi-step execution with the right guardrails and human oversight.
• Engineer and version prompts and structured outputs, and tune them against real evaluation data.
• Build evaluation harnesses that measure accuracy, faithfulness, safety and latency — and use them to ship with confidence.
• Add guardrails, safety and cost controls so systems behave under real-world, adversarial inputs.
• Integrate GenAI into enterprise products and workflows, and monitor them live with the team.

What we're looking for

• 2–6 years in software / ML, with strong Python and recent hands-on GenAI work.
• Built real systems with LLM frameworks (LangChain, LlamaIndex or equivalent) and vector databases (pgvector, Pinecone, Weaviate, Qdrant).
• Real intuition for RAG quality — why retrieval fails, and how to make answers grounded and trustworthy.
• Take evaluation seriously — you measure quality instead of eyeballing it.
• Pragmatic builder who can move fast without shipping something brittle.

Bonus points

What we offer

• Competitive compensation aligned with impact
• Work alongside Kaggle Grandmasters and top AI researchers on frontier problems
• Real ownership and surface area — outcomes, not tickets
• Hybrid flexibility and a learning-first culture via TMLC Academy