Participated · Series A · Jul 2026

Founded in 2024, Prime Intellect offers a full-stack platform for AI agent development that bundles compute access, a reinforcement learning framework, and evaluation tools delivered as a modular marketplace. The platform enables customers to pick specific tools without being locked into an all-or-nothing system. Prime Intellect sells a hosted version of its tools and counts customers including Ramp, Zapier, and Flapping Airplanes. The company reports an annualized revenue run rate of $100 million. Prime Intellect positions itself to let enterprises train and refine agentic systems in-house rather than rely on closed frontier AI labs. Ramp reported that an agent built with Prime Intellect outperformed frontier models on accuracy while running faster and at a fraction of the cost.

Participated · Equity · Jun 2026

Founded in 2023 by Anton Gladkoborodov and Oleg Shevlyagin and based in New York, Corca Research builds a collaborative workspace for mathematics that combines WYSIWYG math editing, real-time collaboration, and built-in AI assistance. Its platform’s AI features help solve problems, manipulate expressions, and generate code so users can work without switching tools. The company emphasizes making mathematical work faster and computationally connected through integrated tooling. The recent $7.8M financing is intended to accelerate product development, bolster AI capabilities, and expand the engineering team. The article does not disclose revenue, users, or other financial operating metrics. Corca has institutional backers including NEA, Bloomberg Beta, Daft Capital, and NVentures.

Participated · Seed · May 2026

Inherent is a London-based AI lab founded by former DeepMind, Microsoft, Reka AI and White House employees. The company is building an AI system called Faraday that it says allows humans and self-improving AI to collaborate on difficult scientific problems. Inherent has framed its mission as writing the playbook for "AI-native science" and reimagining the scientific method from first principles. The founding team includes Tantum Collins, Edward Hughes, Louis Kirsch and Kaloyan Aleksiev, and Matt Clifford is listed as an advisor. Public reporting does not provide revenue, user, or other operating metrics.

Participated · Series B · May 2026

Orbital Industries uses an “AI industrial” approach combining materials discovery, engineering, and manufacturing to accelerate physical product development. Its simulation engine, Orb, models quantum mechanical behaviour of atoms at scale to speed materials simulations and product iterations. Commercially, the company is targeting data centre infrastructure via Orbital IT, developing a PFAS-free dielectric cooling fluid and a refrigeration system for high-density GPUs as well as modular off-site manufactured data centre units for faster deployment. The company was founded by Jonathan Godwin, James Gin-Pollock, and Daniel Miodovnik and is based in London. Financially, it has just raised $50 million in a Series B to scale products, grow its AI and engineering teams, and expand industrial applications beyond data-centre use cases.

Participated · Seed · May 2026

Tensormesh builds and sells Tensormesh Inference, a SaaS inference platform that uses KV caching to store and reuse computed model state to avoid redundant GPU computation. The platform delivers up to 10x reductions in latency and GPU spend by serving repeated context from cache and reports cache hit rates, token-level cost breakdowns, and other performance metrics in real time. Tensormesh offers serverless, OpenAI-compatible API access for immediate use and reserved deployments for enterprises needing dedicated capacity and SLAs. The company also maintains and contributes to the open-source LMCache project (over 8,000 GitHub stars) and highlights integrations across vLLM, SGLang, TensorRT, and multiple cloud vendor offerings. Tensormesh is led by CEO Junchen Jiang and was founded by faculty, PhD researchers, and alumni from the University of Chicago, UC Berkeley, and Carnegie Mellon. The company has raised $24.5 million in total funding and markets a pricing model that bills cached input tokens at $0 for serverless deployments, with well-optimized customers regularly achieving cache hit rates above 70%.