ML Engineer, Retrieval & Grounded Generation

Role overview

Build production retrieval and grounded-generation capabilities for a platform that brings together records from many sources and helps human reviewers understand what the evidence supports. The work spans the full model-serving lifecycle, with a focus on scalable retrieval, verifiable citations, clear operational telemetry, and reliable operation in a secure cloud environment.

Responsibilities

- Implement embeddings, vector storage, and retrieval over a large evidence base with provenance tracking. - Integrate language models so generated text is tied to cited source records, and test for unsupported or failed citations. - Design prompts and structured output schemas for model-assisted workflows. - Package, version, serve, and roll back models; instrument retrieval and generation quality, latency, throughput, overrides, abstentions, and grounding failures. - Use bounded model assistance for difficult narrative extraction, preserving source passages and the exact rule context supplied to each step. - Maintain an additional serving path within the secure environment so model operations are not dependent on one inference provider.

Requirements

- At least five years of experience, including building a production or near-production retrieval-augmented generation system. - Strong Python skills and hands-on experience with embeddings and vector retrieval at scale. - Ability to explain retrieval design choices, grounding evaluation, real citation failure modes, and rollback procedures. - Clear experience distinguishing deployed work from prototypes or proposals. - U.S. citizenship and an active Secret clearance required at the start of employment.

Nice to have

- Deploying models in restricted or disconnected environments, or operating self-hosted and open-weight models. - Fine-tuning, adapters, or custom embeddings. - Federal DevSecOps, risk-management, authorization, or defense cloud experience. - An active Top Secret clearance.

Benefits and work setup

The role is fully remote with occasional travel to offices, customer locations, or partner facilities. The listed compensation range is $165,000–$200,000, with potential bonus and equity. Benefits include employer-paid medical, dental, and vision coverage for employees and their families, manager-approved unlimited paid time off, flexible scheduling, and 14 weeks of paid parental leave.