Participated · Equity · Jun 2026

SE3 Labs builds a spatial AI platform that combines proprietary hardware, AI, and natural language interfaces to help autonomous drones, unmanned ground vehicles, and robotic swarms perceive and navigate complex environments. The company targets defense, public safety, and industrial applications and positions its platform to improve decision-making for those clients. SE3 Labs emerged from stealth in 2026 and is headquartered in Munich. Founders include CEO Lukas Kostler, Chief Science Officer Daniel Cremers, and CTO Simon Klenk. The company announced a €5.5M financing round to fund continued platform development and scaling for its target markets. No revenue or user metrics were disclosed in the announcement.

Participated · Seed · Apr 2026

Auryx builds a software platform that analyzes acoustic signals from the heart, lungs, and blood flow using microphones embedded in everyday consumer hardware. The company’s approach focuses initially on earbuds, leveraging the relative stability of the ear to capture cleaner physiological signals than wrist-based optical sensors. Auryx’s acoustic sensing can extract metrics beyond heart rate, including respiratory rate, cardiac output, and blood pressure according to the company. The founding team comprises researchers from the University of Cambridge, including CEO Erika Bondareva, CTO Kayla-Jade Butkow, and Chief Scientific Officer Professor Cecilia Mascolo. Auryx recently raised $2 million in pre-seed funding to accelerate hardware integration, model development, commercial partnerships, and team expansion. The company plans to expand from earbuds to other microphone-equipped devices and additional health applications.

Participated · Seed · Apr 2026

ManaMind develops autonomous AI agents that continuously play through games to detect bugs, generate actionable reports, and automate repetitive QA tasks. The company has built proprietary visual models tailored for virtual environments and positions gaming as its launch market. ManaMind has formed design partnerships with Included Games and Crazy Labs, indicating early commercial traction. The founders are Emil Kostadinov and Sabtain Ahmad and the company is based in London. ManaMind plans to expand its technical team and accelerate model development with proceeds from its $1.5M pre-seed round. Over time the company intends to broaden its autonomous testing layer beyond games into other software and robotics applications.

Participated · Seed · Mar 2026

deeplify develops an end-to-end AI platform that connects workflows from raw sensor data to automated defect analysis and auditable reporting for industrial inspection and asset integrity management. The platform aims to replace fragmented, manual workflows—such as spreadsheets, static documents, analogue imagery, and manual reporting—with a unified, digital-first system to reduce inspection time, minimise errors, and improve traceability. The solution is grounded in real-world industrial experience from early projects that revealed significant inefficiencies. deeplify completed an initial deployment with Open Grid Europe and ran pilots with SKF, and its platform is used by inspection firms serving global energy companies including Shell. The company is based in Germany. The newly raised €2.0M pre-seed round will be used to expand platform capabilities and accelerate sector deployments across energy, oil and gas, chemicals, and transportation.

Participated · Seed · Feb 2026

Founded in 2023 by Bragadeesh Suresh Babu and Lannan Jiang, Tattvam AI is building an AI system that deeply understands circuit structures and autonomously solves complex semiconductor design tasks. Its reasoning model works from first principles, factoring in constraints, trade-offs, and interdependencies to cut typical chip design cycles from years to weeks. By automating key elements of electronic design automation (EDA) workflows, the company seeks to make custom silicon more accessible, lower development costs, and enable rapid iteration on next-generation chips. Tattvam AI’s technology targets purpose-built processors that can deliver up to 100× performance gains over general-purpose GPUs for specific workloads such as AI training and inference. The startup plans to launch its first commercial product in the coming months while collaborating with leading chip design teams. Proceeds from its recent financing will be deployed to grow the engineering team, accelerate research, and expand industry partnerships. No revenue or user metrics have been disclosed to date.