AI Engineer – Build Legal LLM Agents

About the role

Wexler is creating a next‑generation AI platform for litigation, helping top law firms extract critical facts from massive document sets. As an AI Engineer you will lead the design and deployment of production‑grade large language model (LLM) agents that power this platform.

Key responsibilities

• Design, build, and iterate on LLM‑based agents and proprietary retrieval‑augmented generation (RAG) pipelines.
• Develop robust evaluation frameworks, benchmark models on large legal datasets, and define performance metrics.
• Drive advanced prompt‑engineering practices to maximise model efficacy while controlling token usage, cost, and latency.
• Implement core NLP components such as tokenisation, embeddings, and bias mitigation, ensuring outputs remain structured and cost‑efficient.
• Maintain clean, modular Python back‑end code for data‑intensive systems.

Required profile

• Proven experience shipping production LLM systems.
• Hands‑on expertise designing LLM agents and RAG pipelines.
• Strong background in NLP fundamentals and AI evaluation pipelines.
• Ability to balance real‑world constraints (token limits, cost, latency) while delivering reliable, structured outputs.

Required skills

• Python (expert level)
• LLM agents development
• Retrieval‑augmented generation (RAG) pipelines
• AI evaluation frameworks
• NLP fundamentals (tokenisation, embeddings, bias handling)
• Prompt engineering