AI Engineer

About us 

Biology is going through a revolution. Exponentially increasing data generation, laboratory automation and new AI/ML techniques are reorganizing the industry around computation. At Latch, we've spent the last four years building data infrastructure to store, visualize and analyze the deluge of data in this new era of engineering biology. 10K+ scientists and 7/10 top market cap pharmaceutical companies rely on our product to make medicines.

In the past few months, agentic systems that allow scientists to process their own data with natural language requests - from raw machine outputs to publication ready plots - are actually starting to work. We believe this will soon be the standard way biologists interact with data. But numerous engineering and scientific challenges await in building AI tools that are well scoped, reliable and produce economically valuable work in this domain.

Our tiny team (13 as of July 18th, 2025) has over four years built and sold a large amount of genuinely useful, interesting tech to a difficult market which is skeptical of newcomers. Our customers have a broad range of product expectations, require close collaboration and nuanced communication, include both technical and non-technical users, and routinely execute computational tasks involving multiple terabytes of data. We're beyond excited to add an agentic layer to the existing BD + product machine that already achieved 4.5x YoY growth to date.

About the role

Here are the ideal skill categories:

- Understanding of biology: details of new measurement tech (spatial, single cell) and intuition for appropriate scientific results
- Engineering ability: high quality code, ‘doing things well’
- Product intuition: distill user data into sensible requirements; good intuition for scientist usability
- LLM engineering: understanding of where the 'levers' are with these systems to improve performance; well read on latest practical techniques

Here is a sample of the sort of tech you and your agents (as tools) will interact with:

- Distributed file-system with metadata on Postgres and blobs on S3. Includes a web UI, and a FUSE driver
- Workflow orchestrator based on Kubernetes
- Interactive on-demand compute instances based on Kubernetes containers
- Statically-typed tabular data storage engine
- Reactive Python-based web application framework for data analysis and visualization

Here are some hard technical requirements:

- Strong foundation in statistics, linear algebra, and algorithmic reasoning, with an understanding of the methods underlying bioinformatics tools and LLMs. 
- Understanding of mechanics of wet lab biology, either from direct experience or interfacing with scientists on real projects for some time
- Reasonable software engineering ability - exposure across the stack, experience with databases, understanding where and why performance breaks down under load of data

How we work

- 5-day weeks, in-person in San Francisco Mission Bay
- 1-week-long sprints
- Deep collaboration between Engineering, Bioinformatics, Product, and Sales
- Engineers attend weekly: only 2 company-wide meetings (planning and demo) + 15-min stand-up each day
- Tech: TypeScript, Python, React, Postgres, Docker, Linux, Kubernetes. Lots of one-offs in Rust, Go, others.
- No dedicated frontend/backend/DBA/etc. roles

What we offer

- Good salaries and high equity offers
- Best benefits we can find
- A very nice office
- All meals provided
- Extremely flexible time off
- Snacks, toys, office dog