Exa is building the search engine for the age of AI — from the silicon up. We run one of the most ambitious indexing operations in the world: crawling the open web at massive scale, training state-of-the-art embedding models to understand it, and powering everything through our own high-performance Rust-based vector database and a $5M H200 GPU cluster that regularly lights up tens of thousands of machines.
To do that well, we need to make sure every part of the company is running as efficiently as possible. In our world, there’s a huge difference between 100ms and 200ms. We care about the tiniest details that nobody else does. That’s where you come in.
We are looking for an Automation Engineer to creatively look at company problems and find where anything (code, processes, the coffee order) can be optimized - and then doing it. The role involves being passionate about AI and using it to radically accelerate how we operate. You’re likely a power user of cutting edge AI tools and a forward-thinker about how technology integrates into our lives. You’ll come up with ideas and workflows that we can’t even imagine until it’s obvious in retrospect.
Desired Experience
Strong engineering fundamentals and can build reliable systems
Passion and deep expertise around coding agents like Cursor, Claude Code, Devin
Allergic to technical debt and redundancy
Creative thinker in engineering and operations. Doesn’t accept the status quo of traditional processes
Example Projects
Sit with each team, figure out inefficiencies and how to automate them
Partner closely with ML researchers, infra engineers, full-stack engineers, product, and operations to cut out bloat and enable lightning-fast execution speed
This is an in-person opportunity in San Francisco. We're happy to sponsor international candidates (e.g., STEM OPT, OPT, H1B, O1, E3). In addition to premium healthcare benefits (medical, dental, vision), we also offer fertility benefits and a monthly wellness stipend to all of our employees.
Top Skills
What We Do
Exa was built with a simple goal — to organize all knowledge. After several years of heads-down research, we developed novel representation learning techniques and crawling infrastructure so that LLMs can intelligently find relevant information.
