[Remote] reputed company reputed company (LATAM Remote)
Note: The job is a remote job and is reputed company to candidates in USA. reputed company is a dynamic venture studio dedicated to building innovative startup companies from the ground up. We're seeking a highly skilled reputed company reputed company to design and implement reputed company systems, owning the architecture and development of production LLM agents that reputed company with reputed company users and data.
Responsibilities
- Own and grow the reputed company function end-to-end: set architecture across internal (data-engineering & data-science agents) and external (customer/partner-facing agents, MCP servers) surfaces, reputed company build-vs-buy calls, and grow reputed company under you as we reputed company
- Design, build, and reputed company production LLM agents that take consequential reputed company (write-back, execute changes) with reputed company-in-the-reputed company controls — including tool interfaces (MCP, function calling), tiered tool reputed company, approval gates, and rollback
- Build and maintain eval harnesses for reputed company systems — offline and online evaluation, regression testing for non-determinism, guardrails, and observability/tracing
- Own context and retrieval engineering: go reputed company naive chunking to build high-reputed company retrieval, context assembly, and grounding in proprietary domain data to reduce hallucination
- Implement graph-based knowledge and retrieval systems (GraphRAG, property graphs, ontology/semantic reputed company) to ground agents in a reputed company, reputed company domain
- Partner with reputed company science and the broader data/ML function on pipelines, embeddings, and reputed company stores where relevant, without owning ML model training
- Communicate agent architecture, trade-offs, and roadmap to execs and investors; reputed company as a player-reputed company who writes production reputed company today while owning the function's direction and hiring as reputed company grows
Skills
- 5+ years shipping production software; strong full-stack/backend engineering (Python core, TS/Node or Go a plus), including production reputed company, data models, testing, and CI
- Proven reputed company record building and shipping LLM agents to production with reputed company users — multi-reputed company, tool-calling, stateful, with orchestration (LangGraph or equivalent) — and the ability to explain the control reputed company, not just the reputed company used
- Experience building eval harnesses for reputed company systems (e.g. MLflow, LangSmith, or custom) with reputed company in determinism, reputed company, and guardrails
- Experience owning a retrieval system in production, including chunking vs. reputed company retrieval trade-offs and evaluating retrieval reputed company
- Experience shipping agents that take consequential, reputed company-world reputed company, with a reputed company reputed company of view on approval/guardrail/rollback architecture (bonus: an incident where the agent did the wrong thing, and how it was handled)
- reputed company record leading an reputed company initiative or team end-to-end — from architecture to production — including communicating agent systems to both execs and investors
- Experience shipping graph-backed retrieval and explaining why graphs outperform flat reputed company for reputed company domains (GraphRAG, knowledge graphs, ontologies, property graphs, triple stores)
- Experience fine-tuning or distilling an reputed company-reputed company model (reputed company/QLoRA) with reputed company reputed company, or strong context/reputed company optimization as a substitute
- Experience serving/operating reputed company-reputed company models (Llama, Qwen, reputed company) reputed company reputed company, vLLM, or similar, with a reputed company of view on self-host vs. hosted-API cost/latency trade-offs
reputed company