Anthrogen develops protein foundation models to develop new-to-nature peptides with applications from human health to frontier manufacturing efforts. They are comprised of scientists from Columbia, Duke, Berkeley, and more.

They have recently trained a large, multimodal foundation model that is SOTA across a variety of benchmarks and can even be prompted with natural language. They continue to run scaling experiments and are working towards a public launch of their first-generation models in the near future.

The team is seed-stage and has raised $4M from investors like Paul Graham, Box Group, Y Combinator, Wayfinder, and others. www.anthrogen.com

Skills: Machine Learning

Responsibilities

Work on in-house foundation protein models — Design, Build, and Test
Work with large biological datasets and clean + understand them.
Work with lab team to produce the best proteins possible
You will be expected to be deep in the code and propose ideas
- We will NOT be hiring any “ML engineers” or “Idea people”
Work with the team on all of the above and more!

Qualifications

PhD in Computer Science/Statistics/Machine Learning/Applied Math.
- Open to extremely qualified MS
Worked with unstructured data — EMG, Image, Video, Protein etc.
Sound fundamental understanding of Machine Learning
Proven ability to apply Machine Learning skill — publications, projects, etc.
Strong math + stats background.
Must know how to code well
Has led/initiated their own research processes and is comfortable in a fast-paced environment.
Will learn and adapt quickly—we are an early-stage startup and things will break. We want someone who is willing to roll their sleeves up with us and fix things.
Excited about what we do! We are highly mission driven if you don’t care about making an impact we are not the right fit. If you aren’t willing to do everything it takes to change the world we are also not the right fit. Of all the things on this list, this is perhaps the only non-negotiable.

Benefits

Competitive pay ($120k-180k/year) + equity options (0.5-1.5%).
Full health and dental.
We’re based in SF. Depending on the nature of the work, if relocation is necessary, we will make sure to cover all related costs.

Interviewing for a Machine Learning Research Scientist at Anthrogen

Thank you for taking the time to learn more about Anthrogen’s interview process.

Here at Anthrogen, we’re looking for research scientists that have a strong fundamental machine learning background that can adapt quickly to different datasets and propose novel model architectures. We are looking for people who have a strong sense of ownership and autonomy, as our small research teams largely define their own roadmaps

Our interview process is designed to find these kinds of scientists. Moreover, we try to ask questions that are relevant to the types of models we build -- mostly fundamental machine learning questions. We are a startup at the end of the day so we must be adaptable

Below is what to expect in our interview process. If you have questions, email us at [email protected]

Introduction Call

This stage is a more informal call to get to know more about the role, and for you to evaluate whether it’s a good fit for you. We are looking for people with a strong interest in startups and and the technical background to be able to accelerate our model development; this is usually exhibited by past experience publishing papers or working at other corporate research labs.

We are a small research team that will scale quickly in the next year. There will be many opportunities to grow and potentially manage. We are looking for people who have many good ideas and love to test the ideas themselves. In this chat, you’re welcome to ask lots of questions about whether working at Anthrogen makes sense for you.

Lastly, the conversation will cover some light technical discussion about Machine Learning and your past projects. Mostly, we want to know that you can code and ship fairly involved systems, and speak to them with clarity on the details. This tends to be a good indication that you had strong ownership and showed autonomy in making product or technical decisions.

Culture Fit

Our culture fit chat (~1-1.5hr) will be conducted in-person (virtual can be accommodated). We will be assessing how much you like us as a team and how much we vibe with you. We want to make sure that we will work together well through hard times, and that you are the type of person that is willing to make sacrifices to make this thing work.

We are a very mission driven team and everyone is expected to make sacrifices, but we will never ask anything of you that we would not do ourselves.

In this chat you are encouraged to ask questions about us as people and should be expected to be asked similar types of questions. Feel free to ask additional questions about working environment day to day, anything that is non-technical.

Technical Take home

The first technical assessment is a take home paper with some of the pressing issues that we have recently solved or are currently trying to solve and will be followed by a 75min interview. You will be expected to spend ~2-4hrs on the take home to solve the problems and prepare a presentation on your solution.

The interview comprises of a ~25min presentation of one of you previous publications, projects, etc. followed by the presentation you have prepared for the take home.

The first presentation should be one that you have presented before, ideally one that you are very familiar with and have presented many times in the past — be prepared to answer questions. The next part of the technical will be the ~25min presentation of the take home paper. We will be asking questions about you ideas and design choices.

Feel free to use any resources, this take home is completely open book.

We will compensate for the expected time spend of the take home

Onsite Interview

The interview process is a deeper dive into technical abilities, product thinking and overall fit. For the “fit” portion, we are again looking for ownership and autonomy, but we also dig deeper into communication. This is evaluated throughout the process -- in how you communicate your approach, solutions and overall thinking.

We will have a project prepared and you have the day to try complete it. We will be evaluating you thought process working style, and ability to dig deep to get things done. During this we will be here to assist as if we were your team.

We have found that going through this process gives us enough to make a go/no-go decision. We offer feedback to candidates who request it, and relate the feedback to the skills, qualification and interview process above. This final interview also gives you a good sense of how we work as a team because you will be working as one of us on a brand new problem for a day.

We will compensate for the your time.

Timing

The typical interview process takes 2-3 weeks from start to finish, but we have accommodated as short as one week. We are happy to accommodate candidates who might have a short timeline or competing offers. The more information you can share about your situation and where you are in the process, the more we can either attempt to expedite/match, or save you time if we cannot.