Ideally 1-3 YoE.

About Us

We build training data to train the next generation of Vision Language Actions (VLA) and World Foundation Models (WFMs). It is one of the hardest open problems in physical AI.

What You Will Do

You will build hardware systems fast. You will work on real, messy, failure-prone systems that need to work reliably. You’ll pick sensors, read datasheets, wire things up, break them, and fix them repeatedly until they work.

You’ll integrate sensors with microcontrollers and single-board computers, bring up systems from scratch, and debug across hardware and software boundaries. Expect to move quickly from idea to prototype, iterate under pressure, and make pragmatic engineering decisions. You’ll also handle basic 3D design when needed to make systems usable in the real world.

This role is about ownership, where you don’t wait for instructions, you figure things out and ship working hardware.

Who We Want

We’re looking for builders, who should be comfortable getting your hands dirty - reading datasheets,



debugging hardware issues, and working through ambiguity without step-by-step guidance. You’ve likely built things on your own (robots, IoT systems, embedded projects, anything real), and you understand how components come together beyond theory. You don’t give up when things don’t work the first time (they won’t).

You can move quick, learn faster, and care about making things actually work and not just look good on paper.

Bonus if you’ve worked with embedded systems, sensors, or done any kind of full-stack or ML-adjacent work - but the most important thing is that you can build and figure things out independently.

Location: Bengaluru

Joining: First week of Sep

- To apply, share your resume with a 3 line description of best project you have worked on at [email protected] along with your college name in the subject.

📌 Robotics Engineer (Bengaluru)
🏢 Saturn Labs
📍 Bengaluru