资深软件工程师，代码强化学习

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.

Code RL at Anthropic drives reinforcement learning efforts behind Claude's coding capabilities, creating and scaling agentic coding environments. This is an engineering role with unusual latitude to set technical direction and standards.

You'll be part of a team solving the engineering side of research efforts such as embedding with research teams, getting up to speed on their systems and needs, and designing the frameworks, APIs, and infrastructure that let researchers move faster, then rotating off, leaving behind well-oiled systems those teams can understand, own, and maintain themselves. Your remit also includes the ongoing health of production RL runs: maintainable, monitored, and straightforward to triage.

The team's problem space spans the client side of sandboxed execution for agentic RL environments, large-scale data processing jobs, the lifecycle of production datasets, and the frameworks researchers build environments on. You won't own all of this yourself, you'll take on the slices where your depth matters most. You'll be a strong fit if you have deep expertise in Python, a refined sense of taste for API and framework design, and hard-won intuition for how complex systems fail — especially silently. You should be comfortable diving into messy research code, finding the load-bearing abstractions, and improving them incrementally while researchers continue to build on top of your work.

• Design widely-used APIs, frameworks, and abstractions that other engineers and researchers build on, with careful attention to interface legibility and principled defaults

• Embed with research teams on a rotational basis: understand their engineering needs, build systems and APIs that support their work, and transfer ownership so teams can maintain those systems after you rotate off

• Work directly in research codebases, improving reliability and structure without slowing down the research they support

• Anticipate silent failure modes and prevent them structurally through type safety, well-designed invariants, targeted testing, and refactors that shrink the surface area for bugs

• Contribute to the reliability of production RL systems, including monitoring, regression detection, and triage tooling

• Help define engineering standards, review practices, and design patterns for a new team, and mentor researchers and engineers in adopting them

• Deep expertise in Python, including static typing, safe async and concurrency patterns, and writing performant Python code

• A track record of designing intuitive, safe APIs or frameworks that other engineers or teams adopted and built on

• Experience working productively in large, evolving, or research-style codebases that you didn't originally write

• Demonstrated ability to anticipate failure modes — especially silent ones — and prevent them structurally through system design, type safety, and testing

• Strong written and verbal communication skills, including the ability to explain system designs to collaborators with varied engineering backgrounds

• Comfort with ambiguity: able to scope your own work from a loosely defined problem and drive it to a maintainable outcome

• Experience building infrastructure, tooling, or frameworks for machine learning research or RL workflows

• Familiarity with reinforcement learning concepts, agentic systems, or LLM training pipelines

• Experience building client libraries or SDKs on top of sandboxed, containerized, or remote execution platforms

• Experience designing plugin systems or extensible class hierarchies used across an organization

• Experience embedding with or consulting for other teams, including successfully handing off systems for others to own

• Experience defining code standards, lint rules, or static verification approaches adopted across multiple teams

• Prior experience as a technical lead, or setting engineering standards for a team

• Prior experience maintaining an open source project

Representative projects

These are examples of the challenges the team tackles; no one person will work on all of them:

• Design a base RL environment abstraction general enough to be subclassed across a wide range of environments

• Design a model-tool interface for sandboxed agentic environments that has explicit serialization semantics

• Partner with the platform teams that own the sandbox runtime to specify low-level features that improve the integrity of agentic coding tasks

• Design probes that catch sandbox regressions early

• Design the lifecycle and maintenance scheme for a production dataset

• Lead a research code refactor replacing loosely structured data containers with equivalents that carry stronger correctness guarantees, without breaking the experiments that depend on them

• Design lint rules and code-style requirements that favor statically verifiable patterns — including patterns less likely to be overlooked by an LLM reviewing or writing the code — to shrink the surface area for silent bugs

• Build the access layer that lets researchers discover and reuse data artifacts across teams

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: $405,000 — $625,000 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.