Based out of Bengaluru. Founder of FPV Labs — we build
purpose-built hardware that captures egocentric videos to train the next generation
of vla and world models. Before this, I spent a decade building various startups.
Things I believe in
The data problem in robotics is a hardware problem in disguise.
Human episodes are the right substrate for solving the robotics data gap.
Curiosity is a moat. The people who genuinely want to know how things work tend to outlast the ones optimizing for outputs.
If you are lost, the right thing to do is to build an action-insight flywheel. Action creates insight, which leads to better action and so on. Iterate with every step you take.
The best way to predict the future is to create it.
The best way to learn something new is to build it.
Taste, agency and long term thinking are the best differentiators in an age where superhuman intelligence is available on demand