Participated · Seed · Aug 2026

Datahive AI operates a DePIN-style product that runs on users' phones, turning idle devices into workers that collect public web content such as product pages, reviews, and videos. The company emphasizes privacy, stating it does not collect browser histories, passwords, or private searches and that even the Datahive team cannot trace activity back to individual users. It further describes a security model that shards user-contributed data into tiny fragments distributed across thousands of devices so no single piece is meaningful on its own. Users earn points for participation, and those points can be converted into crypto rewards. Public reporting highlights the company's positioning in the privacy-first and web3 spaces. The company announced a $3.5M Seed raise from several web3 and tech investors.

Participated · Seed · Jul 2026

SkyPilot builds a managed AI compute platform (SkyPilot Platform) atop its open source project to help organizations operate large-scale AI workloads across hyperscalers, neoclouds, Kubernetes clusters, and multiple accelerator types. The platform supports development, agentic workloads, training, reinforcement learning, inference, evaluations, and multi-cluster production serving while offering fleet-wide GPU health monitoring, automated remediation, team and quota management, single sign-on, and SOC 2 compliance. The SkyPilot open source project has surpassed 14 million downloads and attracted over 280 contributors, and the company says hundreds of organizations use the project. Private preview customers have used the platform to manage over 10,000 GPUs and support more than 200 researchers per organization, with some reporting up to 20x performance improvements versus the open source tooling. SkyPilot was founded by Berkeley researchers Zongheng Yang, Zhanghao Wu, and Romil Bhardwaj alongside Ion Stoica and Scott Shenker, and the company launched from stealth alongside its $20M seed raise to fund product, engineering, and go-to-market expansion.

Participated · Seed · Jun 2026

Architect Labs develops an AI-driven system to design and verify custom chips end-to-end, aiming to turn demanding workloads into production-ready silicon. The company plans to expand its capabilities across the full computing stack, including co-designing compilers, runtimes, system software, and ultimately co-optimizing AI models with hardware. Founded by Ebrahim Hussain and Aaditya Subedi, the team includes frontier AI, systems and silicon researchers who collectively have completed 80+ tape-outs and held leadership roles at Intel, Meta, Anthropic, DeepMind, and xAI. Architect Labs is based in Palo Alto, California and emerged from stealth concurrent with its seed raise. Its current financial position includes a $24 million seed round led by Kindred Ventures, which it will use to scale compute, advance AI research, and co-design production silicon with partners.

Participated · Series B · Feb 2026

Opaque Systems, spun out of UC Berkeley's RISELab and headquartered in San Francisco, offers a Confidential AI platform that serves as a trust layer for enterprise AI workloads. The product delivers secure, auditable infrastructure that gives organizations runtime cryptographic proof that data privacy, model weights, and policy compliance are maintained before, during, and after computation. By solving concerns around data leakage and compliance, the platform enables customers to move AI deployments from pilot to production four to five times faster. Current users and partners include ServiceNow, Anthropic, Encore Capital, Accenture, and firms across the technology, financial services, insurance, and healthcare sectors. The company recently launched Opaque Studio, a development environment for building and deploying Confidential AI agents with runtime-verifiable guarantees. Looking ahead, Opaque plans to expand its offerings into post-quantum security, confidential AI training, and sovereign cloud environments. To date, the company has raised $55.5 million in venture funding and is valued at roughly $300 million post-money.

Led · Seed · Jul 2025

AirMDR offers an AI-native SOC that pairs autonomous AI analysts with a 24/7 team of human experts to deliver managed detection and response. The platform autonomously triages alerts in real time and the company says its AI handles 80–90% of tasks typically done by humans; a customer reported AirMDR triages nearly 80% of their security findings. AirMDR provides 200+ out-of-the-box integrations and white-glove onboarding so it can support an entire security stack as a fully managed service. The company is running in production across multiple enterprises and emphasizes affordability for SMBs and enterprises alike. AirMDR plans to invest the new capital in R&D to improve its flagship technology and to scale sales and marketing to meet demand. The company will debut its platform at Black Hat 2025 and was recognized in the 2025 Black Hat Startup Spotlight Competition. AirMDR is an AI-native autonomous MDR provider whose virtual analysts perform 80–90% of the triage, investigation, and response typically handled by human analysts. The platform uses natural language processing and automated playbooks that execute investigations and containment in under five minutes. AirMDR ships 200+ integrations out of the box, covering an estimated 90% of customer integration needs and 40–50% more than typical MDR providers. The company targets SMBs that lack in-house security expertise and aims to reduce SOC costs while improving investigation speed and quality. AirMDR plans to use recent financing to expand R&D and accelerate market entry. Management includes CEO Kumar Saurabh and CTO Srikant Vissamsetti, and Foundation Capital partner Sid Trivedi joined the board; the article frames MDR as a $3B market with rapidly growing adoption.