Hey folks,

My friend is looking for a technical co-founder for an early-stage startup in the inference/infra space.

She has been diving into this over the past few weeks - specifically the memory bottleneck in inference, and more precisely snapshotting GPU memory/state to make model startup near-instant. It’s a pretty niche area with lots of existing research floating around, but it needs someone to actually pull it together and make it production-ready.

They already have a solid client lined up who’s ready to pay (and potentially acquire) once they can deliver a working version.

About her:
2x founder, previously built a venture-backed data center startup, more on the GTM and business side.

About you:
Looking for someone who wants to own the technical side - ideally with relevant experience, or just down to crunch it out and figure it along the way. Ideally someone who can see how foundational this could be for AI infra and has aspirations to be in the startup space for the long term.