Technical Program Manager, RL Environments

About Anthropic

Anthropics 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

Anthropics Reinforcement Learning environments are the foundation of how Claude learns new capabilities. As we scale to massive training runs consuming trillions of tokens we need someone to own the operational health and execution of our RL environments data pipeline.

Youll be deeply embedded with Research Infrastructure and Data Operations teams - not just coordinating across them but making hands-on technical decisions about data quality environment configurations and infrastructure priorities. This role requires both the technical depth to debug yield issues and configure complex ML systems and the program management skills to coordinate across multiple teams during high-stakes production runs.

This is operational technical leadership: youll spend your time monitoring production environment health coordinating in-flight changes during active training runs driving infrastructure migrations and ensuring our environment development keeps pace with our ambitious model training roadmap.

Responsibilities:

• Own capacity planning and execution for major training runs: distribute token allocation targets across environment types track team fulfillment in real-time and coordinate production integration
• Serve as single point of contact for RL environment execution status consolidating visibility across teams and providing unified reporting on capacity fulfillment and blockers
• Coordinate in-flight environment changes during active training runs ensuring source of problem yield are producing at the right pace and quality making technical decisions about configuration updates deployment timing and risk mitigation
• Drive distributed technical initiatives by working hands-on with engineering teams on implementation and validation
• Own mid-run and post-run feedback loops: run retrospectives analyzing environment performance data establish ownership coverage for environment health and feed insights back into roadmap planning
• Ensure quality bar and production pace across problem yield sources working in the weeds on both engineering and science aspects of data generation
• Align environment teams to specific configurations and coordinate deployment of environment updates across the organization
• Maintain operational processes for environment health monitoring issue triage and team coordination during production runs
• Partner with Research leads Infrastructure engineers and Data Operations to identify blockers prioritize competing needs and make technical trade-off decisions

About You

This is not a typical TPM role. Youre likely coming from an ML engineering or RL research background and have developed strong program coordination skills rather than being a traditional TPM trying to learn RL systems. This work is necessarily quite technical - youll need to be in the weeds on both the engineering and science aspects of RL data generation. Generic TPM skills wont be sufficient; you need nuanced understanding of data pipelines and the technical judgment to make real-time decisions during production runs.

You have past experience combining hands-on ML work with technical program leadership. Youre someone who can both debug a data pipeline quality issue and coordinate across five teams to resolve it. You thrive in the chaos of production ML systems where a problem discovered mid-run requires immediate technical judgment and cross-team coordination.

Youre comfortable being in the weeds on technical details - understanding nuances of RL training data environment configurations and infrastructure systems - while maintaining the program-level view needed to coordinate complex initiatives. You build trust with researchers and engineers through demonstrated technical competence not just project management skills.

Qualifications:

• Deep technical understanding of ML training pipelines RLHF systems and large-scale data infrastructure - not just familiarity but hands-on experience with production ML systems
• Experience with RL training data generation environment development or ML data operations at scale
• Nuanced understanding of RL training data characteristics quality metrics and how data issues manifest in model training performance
• Background in ML research or ML engineering before transitioning to technical program management
• Experience with reinforcement learning human feedback systems or AI safety research
• Understanding of data quality frameworks testing methodologies and production validation processes for ML systems
• Proven ability to make technical decisions about data quality system configurations and infrastructure priorities under pressure
• Track record of operational ownership for production ML systems including monitoring incident response and performance optimization
• Experience with operational processes for production ML systems including health monitoring issue triage and incident coordination
• Experience coordinating complex technical initiatives across multiple engineering and research teams
• Demonstrated ability to debug technical issues work hands-on with engineers on implementation details and drive migrations to completion
• Strong technical judgment to balance research experimentation needs with production stability requirements
• Ability to build deep contextual understanding of organization-specific systems and make informed decisions with incomplete information
• Comfort with high-stakes environments where decisions impact millions of dollars in compute spend and model training timelines
• Excellent stakeholder management and ability to influence senior technical staff through competence and consistent delivery
• Track record working in fast-moving AI research organizations

Deadline to apply: None applications will be received on a rolling basis.

Logistics

Education requirements: We require at least a Bachelors degree in a related field or equivalent experience.

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 arent 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 youre interested in this work. We think AI systems like the ones were 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.

How were 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. Were 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 aboutour policyfor using AI in our application process