Staff ML and Agent Engineer on Code Understanding

Sourcegraph is building AI tools to solve the biggest problems in the software industry, particularly as codebases grow and agentic development becomes dominant. The Code Understanding team focuses on surfaces where AI meets developers, including:

• Deep Search, an agentic, multi-step answer engine across an enterprise’s entire codebase.
• Query Assist, which turns natural language queries into Sourcegraph query syntax.
• Smart Hovers, concisely summarizing symbols right where developers need it.
• Guided diff review and APIs that both humans and AI agents rely on daily.

This is a staff-level role for a technical leader who will own the hardest, most ambiguous problems in agent engineering, set standards, and influence direction beyond their immediate team. The role involves:

• Agentic systems: Design and harden multi-step, tool-using agent loops behind current and new agentic experiences, ensuring reliability, observability, and affordability at enterprise scale.
• Pragmatic use of evaluations: Craft evaluations to determine when changes actually improve products, avoiding noise and ensuring fast, confident shipping.
• Models: selection, upgrading, and training: Decide which models to run, drive upgrades, and fine-tune models when necessary.
• Retrieval and context engineering: Improve grounding of models in customer codebases through retrieval, ranking, context windows, and citations.
• Cost and latency: Treat cost and latency as product features, optimizing models for economic sustainability.

You’ll work on a small, senior-leaning team that ships quickly, owns product surfaces, and has streamlined product management. You’ll have real agency over technical direction and a direct line to the impact of your work.

Key Responsibilities:

Within one month:

• Get the Code Understanding products and their model/agent pipelines running end-to-end locally, and land your first improvements to a model, prompt, retrieval path, or evaluation.
• Build a clear picture of where AI engineering pain points are and the product surfaces most constrained by them.
• Get to know the team and customers, forming opinions about where agentic products should go next.
• Join the team’s on-call support rotation.

Within three months:

• Own a meaningful agentic slice of the product end-to-end, driving it from problem-framing through rollout and measurement.
• Establish how the team ships model and prompt changes responsibly, including evals, dashboards, and guardrails.
• Begin up-leveling teammates in building with models by pairing, reviewing code, and modeling good agent engineering instincts.

Within six months:

• Become the recognized technical authority for agent engineering and agentic systems on Code Understanding, with teammates and the wider department deferring to your decisions.
• Measurably improve products with better answer quality, lower cost/latency, or new agentic capabilities.
• Set the direction of the team’s roadmap where it intersects agents, backed by evidence and conviction.

About You:

You are a staff engineer and technical leader with hard-won skills across production machine learning, evaluation, and agent systems. This role requires:

• Ownership of a production model lifecycle, including training, fine-tuning, evaluation, rollout, and monitoring.
• Fluency in building agents, designing multi-step agentic systems, and making them reliable, observable, and cost-bounded.
• Strong evaluation judgment, including building representative datasets, meaningful baselines, and useful error taxonomies.
• Pragmatic treatment of cost and latency as product constraints, making measured tradeoffs between quality, latency, and cost.
• Autonomy on ambiguous problems, owning high-technical-risk projects end-to-end.
• Contribution beyond your domain, translating between engineering goals and business objectives.
• Mentorship and up-leveling of teammates, especially in agent engineering.
• Customer and product-driven mindset, comfortable on customer calls and translating feedback into requirements.
• Pragmatism, shipping the smallest correct thing, and maintaining a high-quality bar with simplicity.

Engineering fundamentals:

• Strong software engineering skills to ship production services.
• Comfort across Sourcegraph’s stack: Go on the backend, TypeScript on the frontend, GraphQL, Postgres, Docker.
• Fluency with agentic coding tools, understanding and owning every line of code they submit.
• Comfort in an async-first, multi-service, fast-paced remote environment.

Nice-to-haves:

• Experience shipping an LLM-powered or agentic developer-facing product.
• Experience fine-tuning, distilling, or training models to meet cost, latency, or quality targets.
• Experience with retrieval, ranking, embeddings, or search relevance.
• Experience working directly with enterprise customers and translating their needs into product requirements.
• Experience mentoring or up-leveling engineers, especially in raising a team’s agent engineering fluency.