Lead AI Engineer, LLM Systems

hermeneutic research

About The Position

Hermeneutic Research Labs, the applied research and engineering unit of hermeneutic research, builds a research intelligence platform for traditional and digital-asset markets. Developed with researchers and traders, the platform turns market, on-chain, macro, and news data into event-driven research that informs trading decisions. The platform is expected to meet rigorous, falsifiable standards and be measured by outcomes. The Lead AI Engineer will own the platform’s multi-agent LLM layer as production infrastructure, improving its speed, cost, reliability, and effectiveness. Research queries move through multiple agent stages, model providers, and self-hosted models. The role begins with hands-on coding and experimentation and is expected to grow into hiring and leading an AI engineering group, with a potential progression to Head of AI Engineering.

Requirements

  • At least 7 years of software or ML engineering experience.
  • At least 3 years shipping LLM-backed systems to real users in production, with ownership of latency, cost, and quality.
  • At least 2 years leading other engineers while continuing to ship software.
  • A track record of building a function from nothing.
  • Strong command of inference economics, including tokens, caching, routing, structured outputs, tool use, and quantisation, with cost-per-request figures for systems operated.
  • Measurement discipline: instrument systems before making changes, evaluate against outcomes rather than model judges, build evaluations that reveal issues missed by informal review, and retire components based on data.
  • Strong Python skills, including async concurrency.
  • Experience with self-hosted serving such as vLLM, adapters and GPU capacity planning, or an end-to-end provider migration.
  • Hands-on depth in vector and graph retrieval, including building and measuring systems, and informed judgment about when retrieval is not the right solution.

Nice To Haves

  • Production experience with fine-tuning and preference optimisation.
  • An informed view on when multi-agent frameworks are over-built.
  • Finance or trading exposure.
  • Experience hiring engineers.

Responsibilities

  • Maintain LLM-layer reliability through latency budgets, concurrency and queueing controls, declared degradation modes, structured-output validation, and provider failover.
  • Build evaluation systems tied to research outcomes, including offline datasets based on real queries and outcome labels, online experiments, regression gates for changes, and quantitatively calibrated model confidence.
  • Manage model routing and costs through per-workload cost attribution, difficulty-based routing, caching, batching, and evaluation of self-hosted serving and GPU capacity economics.
  • Manage model strategy and lifecycle, including pinned model versions, replayable test batteries, rollback, provider diversification, and fine-tuning or adapters when beneficial.
  • Shape agent architecture based on evidence, evaluating and retiring components that do not improve results and improving tool use, memory, and retrieval where they affect research outcomes.
  • Ground and trace model outputs through citations, hallucination controls, audit trails, and clear boundaries for what models may assert versus suggest.
  • Review and mentor engineers, then hire and lead an AI engineering group as the function grows.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service