Participated · Seed · Feb 2026

BeyondMath is a London-based deep-tech firm founded in 2022 by AI veterans Alan Patterson and Darren Garvey to create what it calls the world’s largest foundational physics model. Trained directly on first-principles physics, the platform lets engineering teams generate high-fidelity simulations of aerodynamics, thermal management, and other complex phenomena in minutes instead of hours or days. Early customers span automotive, aerospace, and electronics manufacturers, and the company has formed technical partnerships with NVIDIA and AWS. Management claims the technology can be a “ChatGPT moment for physics,” unlocking faster, more sustainable product design across multiple industries. With fresh capital in hand, BeyondMath plans to double headcount, scale commercial deployments, and expand its customer base across Europe, the United States, and Japan. The company has secured $18.5 million in total seed funding to date to support these growth initiatives.

Participated · Seed · Dec 2025

Matta develops a self-learning computer-vision platform that rapidly adapts to diverse manufacturing lines and performs automated quality inspection, anomaly detection, root-cause diagnosis, and corrective-action recommendations. The system combines hardware, factory integration, AI research, and software, allowing most deployments to become operational within hours and to learn a new line in a matter of days with minimal training data. A central dashboard lets engineers monitor every camera, trace parts, and surface bottlenecks across the factory floor, turning tacit human know-how into scalable, data-driven insights. The technology is general-purpose, working across electronics, automotive, defence, apparel, and both manual and robotic stations, and Matta is already collaborating with OEMs so machines can automatically adjust their own settings. With its “sentient factory” vision, the company aims to deliver increasingly autonomous, end-to-end production. The recent $14 million seed financing provides resources to accelerate product rollout, enhance AI capabilities, build self-serve deployment tools, and expand into key manufacturing regions in Europe and the United States.

Participated · Seed · Nov 2025

Parable has built an intelligence layer that ingests millions of activity data points from a company’s workplace software stack to create a detailed ontology and chronology of work. Its platform establishes a baseline of how time is spent, pinpoints costly inefficiencies, recommends AI-driven automations, and then measures the realized operating leverage. Deployed in private instances, the dashboards help executives answer both “What do we do about AI?” and “How do we know it’s working?” Early pilots have uncovered $80 million in potential cost savings and contributed to a 2.5× increase in one client’s market capitalization. Founded by four second-time entrepreneurs and based in New York City, Parable positions itself as the canonical measurement system that sits alongside every other AI tool in the enterprise stack. The fresh capital will allow the company to scale its observability platform and accelerate go-to-market efforts with large enterprises.

Led · Equity · Oct 2025

California-based SirenOpt is developing PlasmaSens, a manufacturing intelligence system that pairs cold atmospheric plasma sensing with machine-learning analytics to detect defects and material inconsistencies in real time. The platform integrates into both roll-to-roll and piece-to-piece production lines, providing micron-level resolution and actionable feedback that can boost yields, cut waste, and shorten product-development cycles. Initial target markets include battery electrode fabrication, aerospace composites, semiconductors, power-generation components, and automotive parts. The company has seen strong early interest from blue-chip customers and plans to complete its first full inline factory deployments in 2026. To speed commercialization, SirenOpt is expanding its engineering, product, and commercial teams and broadening pilot programs across the US, Europe, and Asia. Financially, the firm just secured a $6.5 million equity round and recently won a $2.4 million grant from the California Energy Commission to tailor PlasmaSens for battery-electrode manufacturing.

Participated · Series B · Aug 2025

Bumper provides an automotive payments platform focused on its PayLater product, which allows drivers to finance vehicle repairs and servicing by spreading costs over several months. The company positions itself as a tool to make car ownership more affordable and keep motorists on the road. Headquartered in Sheffield, Bumper has been steadily scaling across Europe, supported by both equity and credit facilities. Since 2021 it has partnered with Secure Trust Bank Commercial Finance (STB CF), gradually expanding its borrowing capacity to fuel growth. In 2023 the firm also closed an £8 million Series B extension to accelerate product rollout and European expansion. The newly enlarged £60 million STB CF facility gives Bumper significant firepower to onboard more garages and grow its customer base. Management emphasizes that the additional capital will specifically drive wider adoption of PayLater while enabling further investments and potential acquisitions. No revenue, user or profitability metrics were disclosed in the article.