Software Engineer, RL Data

About Anthropic Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems. About the role This is a senior, foundational role on a new team: you'll make architecture decisions the rest of the team builds on, and help shape what we build first. The work is hands-on and varied. Some weeks you'll be deep in pipeline or infrastructure engineering; others you'll be tuning prompts until the output is good, or sitting with a research team that depends on your systems and shipping the fixes they need. We're looking for experienced engineers who own outcomes end-to-end — down to reading transcripts, supporting users, and wrangling vendors. Anthropic's RL Data team builds the systems that produce high-quality reinforcement learning data for Claude: data collection pipelines, human feedback tooling, the execution environments RL tasks run in, and the quality assurance that keeps training data trustworthy at scale. Our goal is to make Claude great at real work — especially the work that matters most, like AI safety research and beneficial deployments of AI. (To be upfront: this is dual-use work — it advances general capabilities too.) Key responsibilities • Own significant parts of our stack end-to-end, from technical architecture through the unglamorous operational work that makes it succeed. • Build data collection pipelines, read the transcripts they produce, and iterate on prompts, evals, and graders until the output is good. • Develop and improve QA frameworks to catch reward hacking and ensure environment quality. • Build interfaces that make collecting human data fast and painless for the people providing it. • Harden execution environments — sandboxing, snapshotting, tool coverage — so tasks hold up at training scale. • Embed with the teams and domain experts who use our systems day-to-day, and work with operations, security, and compliance partners to roll our systems out to new users and vendors. Minimum qualifications • A track record of owning major projects end-to-end in fast-paced, ambiguous environments — for example as a founder or CTO, forward deployed engineer, tech lead, founding engineer at a startup, or creator of a substantial open-source project. • Trusted to run key projects: you lead and inspire others, plan workstreams effectively, collaborate with cross-functional stakeholders, and proactively eliminate or escalate blockers. • Strong software engineering skills in at least one modern programming language — we mostly use Python and TypeScript, but care more that you pick new tools up quickly than that you know our exact stack. Familiarity with Docker, Kubernetes, and common cloud infrastructure is a plus. • Effective use of AI tools in your own day-to-day work. • Care about the societal impacts of your work. Preferred qualifications • Experience with reinforcement learning on LLMs, particularly on the data side: creating evals, environments, rewards, graders, or training data. • Experience helping organizations use AI more effectively, including integrating with third-party tools via APIs, CLIs, and MCP servers. • Strong data engineering skills: pipelines that handle large volumes reliably in production, LLM-powered enrichment steps, and a focus on improving data quality. • Experience shipping user-facing products or internal platforms people love: interviewing users, hunting down friction, measurably improving the experience. • Basic familiarity with AI safety or security research. Representative projects • Take a data collection pipeline from research prototype to a production service that serves many research teams — collection, human validation, grading, and everything in between. • Own the program of developing sandboxed execution environments realistic enough for long-horizon, high-tool-use agentic tasks — and harden them so they behave correctly across millions of rollouts in a frontier training run. • Bring a new data source online — from first conversation with a partner organization to data flowing into production training runs — coordinating with product, security, privacy, legal, and infrastructure teams along the way. • Own the QA layer that decides which tasks make it into Claude's training: automated checks and expert review flows (a busy domain expert should be able to validate a task in under five minutes) that hold up when a frontier model learns to game them. • Cut the time from 'rough task idea' to 'task in a production training run' from days to hours. You'd own the direction: figure out where the bottlenecks actually are, then automate, redesign, or delete the steps in the way. The annual compensation range for this role is listed below. For sales roles, the range provided is the role’s On Target Earnings ("OTE") range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role. Annual Salary: $320,000—$485,000 USD Logistics Minimum education: Bachelor’s degree or an equivalent combination of education, training, and/or experience Required field of study: A field relevant to the role as demonstrated through coursework, training, or professional experience Minimum years of experience: Years of experience required will correlate with the internal job level requirements for the position Location-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices. Visa sponsorship: We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this. We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every single qualification as listed. Research shows that people who identify as being from underrepresented groups are more prone to experiencing imposter syndrome and doubting the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if you're interested in this work. We think AI systems like the ones we're building have enormous social and ethical implications. We think this makes representation even more important, and we strive to include a range of diverse perspectives on our team. Your safety matters to us. To protect yourself from potential scams, remember that Anthropic recruiters only contact you from @anthropic.com email addresses. In some cases, we may partner with vetted recruiting agencies who will identify themselves as working on behalf of Anthropic. Be cautious of emails from other domains. Legitimate Anthropic recruiters will never ask for money, fees, or banking information before your first day. If you're ever unsure about a communication, don't click any links—visit anthropic.com/careers directly for confirmed position openings. How we're different We believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts. And we value impact — advancing our long-term goals of steerable, trustworthy AI — rather than work on smaller and more specific puzzles. We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science. We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time. As such, we greatly value communication skills. The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences. Come work with us! Anthropic is a public benefit corporation headquartered in San Francisco. We offer competitive compensation and benefits, optional equity donation matching, generous vacation and parental leave, flexible working hours, and a lovely office space in which to collaborate with colleagues. Guidance on Candidates' AI Usage: Learn about our policy for using AI in our application process.