Participated · Equity · Jul 2026

ZML develops inference-performance software, notably its newly launched LLM inference server ZML/LLMD, which enables a range of open-source large language models to run across Nvidia, AMD, Google TPU, Apple Metal, Intel Arc and other chips. The company’s earlier public project was an open-source inference-focused ML framework released in 2024 and updated in March; ZML/LLMD itself is not open source but launched as a free product to gather usage data. Founder Steeve Morin, formerly VP of engineering at Zenly, leads a tight team of about 20 people based in Paris. ZML positions itself to reduce vendor lock-in and allow enterprises and cloud providers to use mixed chip fleets for cost and energy efficiency. Its cap table includes prominent founders and researchers such as Solomon Hykes, Hugging Face’s Clément Delangue and Julien Chaumond, and Yann LeCun, signaling strong industry support. The company has raised $20 million from a syndicate of VCs to accelerate product development and future releases, while competing with inference-focused startups like Baseten, Inferact, and RadixArk.

Led · Seed · Mar 2026

xmemory develops a "precision-first" memory layer that structures, cleans, and organizes data at write-time for AI workflows to improve reliability and governance. The approach aims to answer about 95% of critical business and system questions with high precision, leaving a small fraction of queries to standard semantic search. Its platform is currently in closed beta and is led by founder and CEO Alex Petrov. The company raised $4 million in pre-seed funding to support pilots with design partners and further product refinement. Investors in the round include Fly Ventures, Begin Capital, AAL VC, 33East, Inovia Capital, and a group of angel investors and mentors.

Participated · Seed · Mar 2026

Laminar provides observability and debugging tooling specifically for long-running AI agents, capturing every LLM call, tool use, function execution, and browser session recording with a single-line SDK integration. Its Signals feature uses AI to surface failure patterns and anomalies at scale, while its Agent Debugger lets developers rerun an agent from any step preserving prior context. The platform syncs browser session recordings directly with traces so teams can see what an agent was viewing when it made decisions. Since launching in 2025, Laminar counts customers including Browser Use, OpenHands, Rye.com, and Alai, and its SDK is integrated with OpenHands and defaults in Browser Use documentation. The company is led by founders Robert Kim (previously at Palantir and Bloomberg) and Dinmukhamed Mailibay (previously at AWS) and recently completed a $3M seed round to fund product and go-to-market expansion.

Participated · Seed · Apr 2025

TheStage AI develops ANNA, an automated inference-optimization platform that reduces model optimization to hours and makes neural networks more cost-effective. ANNA applies advanced techniques to optimize models and supports various hardware setups and cloud environments. The platform offers pre-optimized models for app developers while also providing granular control for custom neural networks. The startup has already served customers including Recraft.ai and Praktika.ai and has collaborated with Nebius. The company has brought on the Liberman brothers as advisors. TheStage AI was founded by Kirill Solodskih, Azim Kurbanov, Ruslan Aydarkhanov and Max Petriev and is based in Delaware.

Participated · Seed · Sep 2023

Deskree is a Canadian‑Ukrainian no-code tool that provides back-end infrastructure for developers. The company is led by founder Dmytro Grechko. Articles report strong user growth, though no specific metrics were provided. Deskree recently raised $1.5 million to support further user growth and product development. The announcement was shared on LinkedIn and linked to an exclusive interview by Silicon Canals. The team is described as strong and growth‑oriented. No revenue or other financial metrics were disclosed in the article.