Research Engineer, Takeoff Intel

Anthropic; San Francisco (CA); Hybrid; USD 350,000 - 850,000; Full time

Benefits offered by this job

Equity donation matching

Generous vacation and parental leave

Flexible working hours

Lovely office space

Job summary

Anthropic is hiring a Research Engineer for the Takeoff Intel team in San Francisco. You will build evaluation instruments, process large-scale telemetry, and help quantify AI capability growth to inform safety and policy work.

The role spans evals infrastructure, data processing, and analysis tooling with an emphasis on shipping reliable results. We value rapid prototyping, accuracy, and the ability to turn vague questions into running instruments.

Qualifications

• You may be a good fit if you have shipped an evaluation, data product, or research library end to end.
• Prototype fast and are comfortable throwing code away.
• Handle messy, large-volume data without over-engineering.
• Have run experiments on large language models, not just moved their outputs around.
• Can work from a vague question rather than a spec.
• Communicate results clearly and collaborate with researchers.

Responsibilities

• Design, build, and run capability evaluations and measurement instruments at scale
• Build the data and analysis pipelines that turn large volumes of model outputs and telemetry into reliable metrics
• Prototype new instruments fast, validate them, and decide what to keep
• Review and supervise AI-written code as a normal part of the workflow
• Work closely with research scientists on the team and with partner teams to define what's worth measuring
• Contribute to internal write-ups and public reporting

Education

Bachelor’s degree or equivalent

Job description

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 Team

At Anthropic, we are delegating a growing share of AI development to AI systems themselves. Takeoff Intel is the team that measures this recursion from the inside. We're part of the Anthropic Institute. We design evaluations of AI R&D capabilities, build the internal telemetry Anthropic uses to track how much of its own model development is becoming AI-assisted, and develop the quantitative methods that turn those signals into a calibrated picture of where capability growth is heading, so that Anthropic and the wider world have accurate situational awareness on this acceleration.

Our work appears in Anthropic's model system cards (we own the AI R&D capability assessments and adapted Epoch's Capabilities Index to our evals); all the data in When AI Builds Itself comes from our team. Internally, our measurements shape research priorities and safety planning; externally, they contribute to Anthropic's public reporting on the pace of AI progress and to collaborations with third-party evaluators. We're a small team that works closely with pretraining, RL, economics, and policy researchers across the company. If you're passionate about measurement accuracy, and feel urgency about safety and situational awareness, you should consider joining us.

About The Role

As a Research Engineer on Takeoff Intel you'll build and run the evaluation and measurement instruments that make this research possible. This is a generalist role on a small team: you'll work across evals infrastructure, large-scale data processing, and analysis tooling, and you'll prioritize shipping. We build instruments that answer real questions and help set priorities, not dashboards that surface noise. We value working prototypes, rapid iteration, accuracy and good prioritization. We often need to go from a vague research question to a running instrument quickly.

We're hiring at both junior and senior levels.

Responsibilities

• Design, build, and run capability evaluations and measurement instruments at scale
• Build the data and analysis pipelines that turn large volumes of model outputs and telemetry into reliable metrics
• Prototype new instruments fast, validate them, and decide what to keep
• Review and supervise AI-written code as a normal part of the workflow
• Work closely with research scientists on the team and with partner teams to define what's worth measuring
• Contribute to internal write-ups and public reporting

You may be a good fit if you

• Have shipped an evaluation, data product, or research library end to end
• Prototype fast and are comfortable throwing code away
• Handle messy, large-volume data without over-engineering
• Have run experiments on large language models, not just moved their outputs around
• Can work from a vague question rather than a spec
• Communicate results clearly and collaborate closely with the researchers whose questions your instruments answer

Strong candidates may also have

• Built evaluation harnesses or benchmark infrastructure for LLMs
• Experience with large-scale ML or data infrastructure (self-driving, observability, or similar) alongside ML exposure
• Built tools or libraries that other researchers rely on
• A track record of catching what AI-written code gets wrong

Some examples of our work

• Anthropic ECI: our adaptation of Epoch Capabilities Index published in all recent system cards to measure capability acceleration
• AI R&D capability assessments in the Claude system cards
• When AI Builds Itself: all data in the article comes from our team

The annual compensation range for this role is listed below.

Annual Salary

$350,000 - 850,000 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.

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.