ML & Agentic Systems Engineer [IC4]

Who we are Our mission is to bring clarity and control to the world’s most complex codebases. AI is accelerating code creation, but the infrastructure to understand, oversee, and evolve that code hasn’t kept pace. Sourcegraph gives engineering organizations full visibility across their systems, precise context for their agents, and the ability to execute coordinated code changes at scale. As agentic development becomes the dominant engineering paradigm, we provide the context layer teams need to take control of their codebase. With Code Search, Deep Search, MCP, and Agentic Batch Changes, we deliver on that mission today - giving engineering teams and their AI tools the cross-repo context to navigate massive codebases with confidence, and the ability to make changes across hundreds of repositories at once. Companies like Stripe, Reddit, and Leidos rely on Sourcegraph to ship faster and with higher quality. We’re backed by a16z, Sequoia, and Redpoint, and proud to operate as a globally distributed team that values high agency, direct communication, and customer love. If you want to build the infrastructure that lets every engineering team - and every agent they deploy - operate on their codebase with confidence, join us. Hours & location 🌎 While we hire almost anywhere in the world, we have a preference for someone to reside in the following locations for this role. However, if you feel qualified, we welcome you to apply regardless of location. No matter what, working hours must overlap with EST for at least 20 hours/week. Preferred locations: Europe North America Why this job is exciting Sourcegraph is at the forefront of building AI tools to solve the biggest problems in the software industry, problems that only get bigger as codebases grow and as more of the work is done by agents. The Code Understanding team owns the surfaces where that intelligence meets the developer: Deep Search , our agentic, multi-step answer engine across an enterprise’s entire lineup of codebases, Query Assist , turning natural language queries into Sourcegraph query syntax, Smart Hovers , concisely summarizing symbols right where devs need it, guided diff review, and the APIs that both humans and AI agents rely on every day. Most of what makes those surfaces tick is agent engineering : a blend of software engineering, machine learning, and statistics. Agent engineering tells us which model to use and when, how to retrieve and pack context, how to measure answer quality, where to fine-tune or distill a smaller model to cut costs and latency, and how to expand a single LLM call into a reliable multi-step agent. As the staff ML and agent engineer on Code Understanding , you’ll be the technical owner of it: setting the tam’s direction for models, evaluations, and agentic systems, making our products measurably better, faster, and cheaper, and raising the team’s fluency in building with models. This is a staff-level role : we’re hiring a technical leader, not just a strong individual contributor. The primary need is a production ML and evaluation authority who also builds production agent systems. You’ll own the hardest, most ambiguous problems in this space, set standards others follow, and influence direction beyond your immediate team. You’ll get the exhilarating chance to drive the vision on how we can provide the best code understanding experience on the market by combining our deterministic, large-scale systems and AI into experiences never seen before. Concretely, you’ll own work like: Agentic systems. You’ll design and harden the multi-step, tool-using agent loops behind current and new agentic experiences, turning research and experiments into reliable, observable, and affordable products at enterprise scale. Pragmatic use of evaluations. Crafting agentic products means needing to tell when a change actually helped, which is hard when agents keep changing and the product keeps shifting. You’ll bring judgment about where evaluations earn their keep, when to use targeted smoke tests and metrics, and how to avoid noise dressed up as rigor, so we can move fast with confidence. Models: selection, upgrading, and training. You’ll decide which models we run where, drive upgrades, and fine-tune our own when that’s the right call. Retrieval and context engineering. You’ll push on how we ground models in a customer’s code - retrieval, ranking, context windows, citations - to make answers more accurate and verifiable. Cost and latency. Every surface has a per-user economic budget. You’ll treat cost and latency as product features, and profile, distill, cache, and right-size models so we can ship ambitious features sustainably. You’ll do this on a small, senior-leaning team that ships quickly, owns a lot of product surface, and has streamlined product management: engineers here talk to customers, frame the problem, and own it end-to-end. You’ll have real agency over technical direction and a direct line to the impact of your work. 📅 Within one month, you will… 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 eval. Build a clear picture of where the AI engineering pain is, and the product surfaces most constrained by them. Get to know the team and our customers, and start forming your own opinions about where our agentic products should go next. Join the team’s on-call support rotation. 📅 Within three months, you will… 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: the evals, dashboards, and guardrails that make quality and cost regressions visible before customers feel them. Begin up-leveling teammates in building with models by pairing with them, reviewing their code, and modeling good agent engineering instincts. 📅 Within six months, you will… Be the recognized technical authority for agent engineering and agentic systems on Code Understanding. You will be the person teammates, and increasingly the wider department, defer to on model, eval, and agent-design decisions. Have measurably moved the products: better answer quality, lower cost/latency, or new agentic capabilities that weren’t feasible before. Be setting the direction of the team’s roadmap where it intersects agents, bringing conviction, backed by evidence, about which bets are worth making, and pulling other engineers up to execute on them. About you You are a staff engineer and technical leader with hard-won skills across production machine learning, evaluation, and agent systems . This high-leverage role relies on your ability to make sound model and evaluation decisions for a fast-moving product, build the production systems around them, steer technical direction, and be a force multiplier for a talented, product-minded team. You’re equally comfortable reasoning about an eval harness, a fine-tuning run, a retrieval pipeline, and the multi-step agent loop that ties them together, and you make everyone around you better at all of it. You operate at staff scope: you own the most ambiguous, highest-risk problems in your domain, go into whatever codebase a problem requires, set standards and patterns others adopt, and translate fluidly between engineering goals and business objectives. You influence direction beyond your immediate team. You lead through technical excellence and mentorship. You have personally owned a production model lifecycle. You have trained or fine-tuned at least one model and taken it from dataset construction through evaluation, production rollout, and monitoring. You can explain how you chose between training or fine-tuning and prompting or retrieval, how you chose baselines and metrics, what error analysis revealed, and how production evidence affected the next version. You build agents, fluently and opinionatedly. You’ve designed multi-step agentic systems and made them reliable, observable, and cost-bounded. You have a point of view on where agents shine and where deterministic code or human judgment is required. You have strong evaluation judgment. You build representative datasets, meaningful baselines, useful error taxonomies, and release criteria that connect offline measurements to production behavior. You know where rigorous evaluations earn their keep, where lightweight smoke tests or qualitative review are enough, and where a precise-looking metric is misleading. You treat cost and latency as product constraints. You make measured quality, latency, and cost tradeoffs and use the appropriate combination of model selection, prompting, retrieval, caching, distillation, and fine-tuning rather than reaching reflexively for a more complex model. You operate autonomously on ambiguous problems. Given a rough product idea, a few customer quotes, and a Slack thread, you come back with a plan, a prototype, milestones, and a point of view on tradeoffs, without it being pre-scoped. You own high-technical-risk projects end-to-end. You contribute beyond your domain. As a senior IC, you go into whatever part of the codebase a problem requires, recognize issues beyond your immediate area, and translate between engineering goals and business objectives. You up-level the people around you. You mentor by pairing on hard problems, providing substantive design and code reviews, and spreading agent engineering literacy across the team. You see investing in your teammates’ growth as part of the job. You’re customer and product-driven. You’re comfortable on customer calls and in feedback threads, you turn raw signals into requirements, scopes, and milestones, and you push back when feedback would lead the product astray. You’re pragmatic, not a perfectionist. You ship the smallest correct thing, prefer robust solutions over complicated ones, and keep a high-quality bar with simplicity. On the engineering fundamentals: You’re a strong software engineer who can ship production services. You’re comfortable across our stack