The work is hard, the problem is real, and the team is small.
BoolSi compiles hotspots in C, C++, or Rust into custom-generated FPGA accelerators: no HDL, no chip-design team. We train ML models to learn a program's behavior and converge them into exact digital circuits.
That puts us at the intersection of compilers, machine learning, and hardware: three fields that rarely sit in the same room. We're early, well-funded, and shipping toward a private beta in Q3 2026. Everyone here owns a real piece of the stack.
We're based in Boston, backed by F-Prime, Pillar VC, Fifth Quarter Ventures, and Coalition Ventures.
Open roles.
-
→
Compiler Engineer
Own the pipeline that lowers a C, C++, or Rust hotspot into a verified circuit on an FPGA, front-end to bitstream.
-
→
ML Research Engineer
Train differentiable models that discover the exact discrete program behind a set of examples and converge into digital circuits. Exact, not approximate.
Don't see your role? If you work at the boundary of compilers, ML, and hardware, we still want to hear from you.
What we offer.
- Competitive salary and meaningful equity, early enough to matter.
- Health, dental, and vision insurance.
- Flexible PTO.
- On-site or hybrid in Boston, or fully remote, depending on the role.
- A learning budget, because nobody walks in fluent in compilers, ML, and hardware all at once.