Today, software engineers can multiply their productivity by 10x using AI tools like Cursor. Design Verification (DV) engineers cannot. Hardware lacks the open training data that enabled the AI revolution everywhere else. The knowledge required to design and verify our physical world is not publicly accessible; nearly all chip code is closed-source. We envision a future where AI discovers chips beyond human imagination, yet the expertise needed to build today's most efficient domain-specialized accelerators lives exclusively within the minds of a few thousand human experts.
There is no path to automated chip design until someone breaks this data barrier.
Phinity is building the foundational training infrastructure that will teach LLMs how to design, verify, and eventually tape out chips.
We work with the largest frontier model labs to build realistic reinforcement learning environments for chip design: complete RTL design challenges, verification tasks, debugging scenarios, and down the road, P&R and validation.
This is the ground floor of agentic hardware engineering.
The work you do here determines whether chip design enters the AI age, or gets left behind.
We're hiring the best RTL design and DV engineers who have worked on the most ambitious custom chips to shape the future of hardware automation. You will collaborate directly with frontier model researchers to create the environments that teach AI how to design and verify custom chips.
Think you're a fit? Email founders@phinity.ai your resume.
Our founders have helped train the best open-source frontier models in RTL code generation at companies like NVIDIA and are backed by $5.5M in funding from Silicon Valley's top investors + DeepMind's Jeff Dean.
© Phinity Labs, Inc. 2025