Sidharth Giri

Figuring out. MTS @proximalhq

[Sidharth Giri @helmholtzigga]

I joined Proximal wanting to work with a highly skilled team moving incredibly fast. It’s been crazy to see just how far we’ve come in such a short time.

Back then, a lot of things were still slow and manual. Watching us progressively automate that while keeping pace with smarter and smarter models has been quite the journey.

Incredible to see how things have taken shape since the early days.

Still lots to come!

[Proximal @ProximalHQ]

Proximal builds infrastructure that enables AI to improve from real-world experience.

In the past year, we have grown to more than $200M in annualized revenue helping frontier labs and enterprises improve coding agents. Now, we are expanding beyond software engineering.

Recent progress in mathematics demonstrates what happens when models are trained in domains with massive amounts of public data and easy verifiability, making it easy to find weaknesses, generate training tasks, and iterate quickly.

We build infrastructure that enables this feedback loop in other domains: we are excited about a world in which AI systems design targeted drugs, accelerate chip design, and rewrite legacy software that still runs hospitals, power grids, governments, and other critical infrastructure.

Since starting, we have assembled a small team of researchers and engineers and built world-class infrastructure for post-training and synthetic data research to overcome the challenges of scaling manual data curation.

Our team comes from Cursor, Google DeepMind, Meta Superintelligence, Prime Intellect, Citadel, and Jane Street. More than half are former founders.

We’re fortunate to be backed by investors who share that vision, including @generalcatalyst who led our $15M Seed Round at a $300M valuation, as well as @chemistry, @svangel, @dvlamoen and individuals like @LiamFedus, @kevinweil, and @bernhardsson

[Sidharth Giri @helmholtzigga]

@akshitkr Very cool work man!

[Sidharth Giri @helmholtzigga]

@feydemon Almost thought that was an army of wall titans

[Sidharth Giri @helmholtzigga]

@llmluthor Ofcourse ofcourse

[Sidharth Giri @helmholtzigga]

@khoomeik quality >> quantity

[Sidharth Giri @helmholtzigga]

@llmluthor Yep, capable agents do eventually saturate the reward on fair tasks

[Sidharth Giri @helmholtzigga]

@sumanthd17 @rkal4 3d printer 🙂‍↕️

[Sidharth Giri @helmholtzigga]

@CFCforxile Neymar Jr.

[Sidharth Giri @helmholtzigga]

@GammaPathR @scaling01 I think 10 vs 20 is still a big cost gap which won't be covered that quickly

[Sidharth Giri @helmholtzigga]

@pHequals7 Frontier SWE is pretty fair imo

[Proximal @ProximalHQ]

Gemini 4 Argon scores 55.0% on FrontierSWE

The model is close to Fable 5.1 and only outperformed by Opus 5.5 and GPT-6 Astra

[Sidharth Giri @helmholtzigga]

@sydtshah The goat himself

[Sidharth Giri @helmholtzigga]

@roerohan How did I miss this

[Sidharth Giri @helmholtzigga]

@akshitkr @ProximalHQ 🐐

[Sidharth Giri @helmholtzigga]

@av_bobby 🐐

[Sidharth Giri @helmholtzigga]

@roerohan @MatternJustus @brendanigraham @TarushNagpal It was such a vibe

[Sidharth Giri @helmholtzigga]

@brendanigraham Unnecessary violation lmao