Guide Startups

attimet

Fall 2024 · Active

we predict stuff

Location: San Francisco, US

Founded: 2024

Team size: 3

About

We’re building a research lab that puts its ideas to the test in one of the most complex, information-rich environments in the world: the financial markets. Much of trading still depends on hand-crafted signals and intuition. We’re approaching it differently - from first principles. We design systems that learn, adapt, and improve with data. Our infrastructure is built to accelerate research: fast iteration loops, real-time feedback, and direct connection between ideas and outcomes. We’re starting with liquid markets like equities and options - high-dimensional, dynamic systems where traditional pipelines break down. Our goal isn’t just to model them better - it’s to build a platform for experimentation, where every result tightens the loop between theory and practice. If you care about learning systems, clean abstractions, and applying research in the wild - we’re hiring.

Founders

  • Kirthi BanothuFounder

    Got into building things back in 2006 with a website for Dragon Ball Z content. Ex-HFT (Optiver + DRW), built low-latency trading systems a few too many times - developed strategies for futures, cryptos, equities, and options. Had a stint in self-driving leading the onboard sensor and ml libraries team at Argo AI. Skipped half as many grades as my cofounder.

  • Xiaoyu LiFounder

    Ex-quant trader at Optiver specializing in index + equity options market-making. Started and grew a medium-frequency algorithmic trading team. Yale Medical School dropout. Skipped twice as my grades as my cofounder.