Tokens Team Manager

Our missionAnthropic is an AI safety and research company that’s working to build reliable, interpretable, and steerable AI systems. Large, general systems of today can have significant benefits, but can also be unpredictable, unreliable, and opaque. Future systems might be very dangerous if they still have these issues. Our goal is to make progress on them.
How you’ll helpYou’ll lead the “tokens” team, which manages infrastructure related to data and prepares data for our training runs. We believe that there are three key ingredients to making progress in AI: compute, data, and algorithms. Data is often paid the least attention to externally, but we think it can make as much of a difference to the quality of the model as each of the other two. This team has three critical responsibilities that all must be executed well in order for Anthropic to succeed:
1. Science of Data: Once we’ve acquired data, we need to figure out the best ways to use it and make sure we’re in shape for training runs. This work primarily consists of coming up with ideas, doing ML experiments, and following up with careful analysis of the results.
2. Data Availability: Large language models need significant amounts of data to be able to train, and having the necessary number of tokens is essential for improving our language models. The data needs to be high quality and diverse for the performance of the model and also to supply research datasets.
3. Infrastructure: The team maintains data pipeline infrastructure as well as the tokenizer required to rapidly process data on the order of terabytes. This infrastructure is relied upon by other research teams to easily manipulate their data as well as the tokens team to quickly turn around training datasets and do research.

As the team lead, you will...

• Be responsible for delivering as large and high quality of a dataset as possible for our large training runs
• Prioritize the team’s work in improving our tokens and infrastructure, in collaboration with senior engineers and other stakeholders
• Steer cycles and processes of our research in support of the science of data work
• Guide the team through growth and evolving research focus in uncharted territory
• Coach and develop your reports to decide how they would like to advance in their careers and help them do so

You might enjoy this role if you…

• Believe that advanced AI systems could have a transformative effect on the world, and want to help make sure that transformation goes well.
• Have an interest in understanding the role of data in language models, working on large scale systems or orchestrating one-off research tasks.
• Are an experienced manager and enjoy practicing management as a discipline.

Requirements...

• You have at least 1 year of prior engineering management or equivalent experience.
• You can come into our San Francisco office at least 25% of the time.

Important skills...

• People management: Coaching, performance evaluation, mentorship, and career development.
• Project (research) management: Prioritization, communicating across team/org boundaries, and keeping research on track.
• Technical expertise: Keeping a strong understanding of which datasets are valuable to the training, understanding how data works in language model training and being able to maintain good practices on the codebase and infrastructure. We expect this to be a skill that is developed on the job, and will be looking for related experiences such as having research experience, being well-read in LLMs, or having led work that involves rapid hacking and interpreting complex results. Skills in research management will be prioritized over engineering management. 
• Recruiting: Predicting staffing needs, designing interview loops, evaluating candidates, closing candidates.

Annual salary (USD)

• The expected salary range for this position is $300k - $520k.

For this role, we prefer candidates who are able to be in our office more than 25% of the time. We have a preference for candidates who are based in the Bay Area, as most of the team is in San Francisco, though we believe that candidates with skills managing a remote team can be just as effective. 
Compensation and Benefits*Anthropic’s compensation package consists of three elements: salary, equity, and benefits. We are committed to pay fairness and aim for these three elements collectively to be highly competitive with market rates.
Equity - On top of this position's salary (listed above), equity will be a major component of the total compensation. We aim to offer higher-than-average equity compensation for a company of our size, and communicate equity amounts at the time of offer issuance.
Benefits - Benefits we offer include:- Optional equity donation matching at a 3:1 ratio, up to