Lead AI Engineer (LATAM Remote)
RemoteUnited States or Mexico
Job Summary
Own and grow the agentic function end-to-end by setting architecture across internal data-engineering and external customer-facing agent surfaces, making build-vs-buy calls, and leading team growth. Design, build, and deploy production LLM agents that execute consequential actions with human-in-the-loop controls, including tool interfaces, approval gates, and rollback mechanisms. Build and maintain eval harnesses for offline and online evaluation, regression testing, and observability. Own context and retrieval engineering, implementing graph-based knowledge systems like GraphRAG to ground agents in proprietary domain data. Partner with applied science on pipelines and vector stores without owning model training. Communicate agent architecture and roadmap to execs and investors while writing production code.
Required Qualifications
- 5+ years shipping production software
- strong full-stack/backend engineering (Python core)
- Proven track record building and shipping LLM agents to production with real users
- Experience building eval harnesses for agentic systems
- Experience owning a retrieval system in production
- Experience shipping agents that take consequential, real-world action
Desired Qualifications
- strong full-stack/backend engineering (TS/Node or Go)
- the ability to explain the control loop, not just the framework used
- a clear point of view on approval/guardrail/rollback architecture
- Track record leading an agentic initiative or team end-to-end
- communicating agent systems to both execs and investors
- Experience shipping graph-backed retrieval
- Experience fine-tuning or distilling an open-source model (LoRA/QLoRA)
- strong context/prompt optimization as a substitute
- Experience serving/operating open-source models (Llama, Qwen, Mistral) via Databricks, vLLM, or similar
- a point of view on self-host vs. hosted-API cost/latency trade-offs
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.