The Venture Codex

Participated · Seed · Aug 2026

Exclaim Robotics develops mobile robots that autonomously navigate data halls, access rack fronts and perform hands-on maintenance tasks inside live server racks. Its initial scope focuses on narrow, high-frequency interventions such as cleaning and replacing fibre-optic connectors and SFPs, handling copper cabling, and swapping failed drives. The company emphasizes operating safely inside live 800-volt racks and aims to reduce labour-intensive manual maintenance while improving operational efficiency for data centre operators. Founder Helen Oleynikova, who has prior robotics experience at Google, Microsoft, Nvidia, Willow Garage and ETH Zürich, leads the team alongside founding engineer Patrick Pfreundschuh, with two more founding engineers planned to join in 2027. Exclaim has positioned itself alongside startups such as Boost Robotics, Watney Robotics and Gecko Robotics in the data-centre robotics market and has stated goals of accelerating product development using its recent funding.

Led · Seed · Apr 2026

Calibre offers a clinician-led service that connects medical history, daily behaviours and environmental context with diagnostics and causal AI to identify drivers of an individual's health and provide personalised plans. It launches a new category it calls Causal Health Navigation and is positioning itself as a proactive health partner available from £69 a month. The company emerged from stealth and has opened a waitlist for its initial rollout. Clinical services are currently provided through partner DocTap, and Calibre has submitted its own CQC application to deliver care directly. Founders include Alexander Weber (formerly early team and Chief Growth Officer at N26), Ben Levy (experienced software and healthtech executive), and Dr Reinhold Innerhofer (medical doctor and sports scientist). Calibre raised $3.3M in pre-seed funding to support its launch and growth.

Participated · Equity · Apr 2026

Modern Relay provides an operational foundation that lets organisations coordinate people and AI agents from a single canonical representation of company context (people, policies, data, and decisions). It released Omnigraph, an open-source, Git-style graph database designed for agents to branch, propose parallel changes, and merge approved updates back into a canonical graph. The company stresses that its infrastructure runs on customer-owned systems to keep organisational data and governance under the customer's control. Modern Relay was founded in 2025 by Ragnor Comerford and Aaron Goh and is headquartered in San Francisco and Barcelona. The team reports internal usage with employees running multiple agents to draft research, ship code, and handle operations, indicating early product traction. The firm's stated mission is to prevent fragmentation as agents proliferate across departments by providing a shared layer for coordination and governance.

Participated · Series A · Mar 2026

Rivia has built a unified data engine that integrates thousands of heterogeneous clinical trial data files in real time and applies trial-specific scientific logic via a proprietary library of reusable configurations. Its platform harmonizes data into operational review workflows and is launching embedded AI agents—starting with Spark, which converts natural language into publication-grade clinical visualizations—with additional agents for proactive data quality monitoring and oversight. The company says its ontology library compounds with each trial, improving performance over time, and that customers have run global trials on the platform producing measurable operational benefits. Rivia positions its stack as an interoperability layer above incumbent systems like Veeva and Medidata, aimed at reducing manual stitching of sources and accelerating decision-making. Near-term plans include scaling the team across Zurich and Boston and deploying next-generation agents to support adaptive and decentralized trial workflows, with the stated ambition of substantially lowering trial costs and speeding insights.

Led · Seed · Dec 2025

Valinor is an AI-driven biotechnology company building multimodal machine-learning models trained on matched multi-omic and clinical outcome datasets to predict patient response to therapeutics. Its platform enables drug developers to distinguish responders from non-responders, refine target populations, and uncover novel biology that can reveal new indications. The company’s proprietary datasets combine patient-derived omics data with real-world treatment outcomes, providing a rich foundation for predictive modeling across multiple disease areas. Valinor operates on an AI-first, disease-focused approach designed to reduce R&D costs, de-risk drug development, and shorten time to market for life-saving medicines. With fresh capital, the team plans to expand its dataset library, hire additional machine-learning talent, and deepen collaborations with pharmaceutical partners. Although revenue or user figures were not disclosed, the seed financing underscores investor confidence in the platform’s commercial potential. The company is headquartered in San Francisco and is actively seeking collaborations for customized patient-selection models.