Led · Seed · Aug 2026

Founded in 2023 by disputes lawyer Hanna Roos and commercially live in 2024, Aavalynx has developed proprietary AI technology that enables legal teams, law firms and litigation funders to structure, analyse and interrogate dispute data at scale. The platform is designed as a central repository and control panel for dispute information, linking every statement to source documents to minimise hallucinations and support evidence-based decision-making. Aavalynx positions its product to shift organisations from reactive dispute management to earlier, data-driven intervention, with co-development partners such as Vodafone. Early metrics cited by the company indicate an ROI of roughly 30x in saved damages, legal fees and interest, and up to 200x when recovered commercial opportunities are included. After the pre-seed raise, the company plans to accelerate product development, grow its team and expand its global client base.

Participated · Seed · Aug 2025

Axelyf develops next-generation lipid nanoparticle (LNP) delivery systems to improve RNA therapeutic delivery, combining novel ionizable lipids, advanced surface chemistry and AI-informed design. Its AXL technology integrates assets acquired from 76Bio with Axelyf’s proprietary IP and four distinct ionizable lipid series designed for potency, biodegradability and extrahepatic delivery. The company also built ANNA™, a machine-learning model for lipid optimization, and published a white paper showing ANNA outperformed current AI approaches for predicting LNP performance. Axelyf plans to use the new funding to advance development of its proprietary AXL technology and CMC capabilities. In partner studies AXL lipids demonstrated superior in vivo gene editing potency versus the LP-01 benchmark, higher mouse gene editing efficiency at lower doses without signs of liver toxicity, and high potency for gene insertion with reduced immune stimulation. Its lead investigational candidate, AXL-003, targets an undisclosed autoimmune condition using liver-directed delivery; the company maintains an R&D subsidiary, Axelyf ehf, in Iceland.

Led · Seed · Apr 2025

Evolytes is an elementary (K-5) math learning platform built around high-quality game design, universe and world-building, and character-driven narratives. The product integrates math practice into a single adventure rather than digitizing exercises or offering isolated mini-games. The company reports serving more than 160,000 students across 1,400 schools in the U.S. and says its platform improves student outcomes, attitude, and engagement in blended and hybrid learning models. Evolytes was founded in 2017 and is Reykjavík-based. The company plans to scale revenue and impact across Latin America, India, the Middle East, Southeast Asia, and the U.S. according to its recent statements.

Participated · Series B · Oct 2024

Centaur Labs uses a network of over 50,000 vetted experts to label multimodal health and science data for model training, evaluation, monitoring, and feedback. It operates a HIPAA-compliant platform and a mobile app, DiagnosUs, where experts compete to be top annotators to ensure high-quality labels. The company says its expert-sourced approach delivers annotation accuracy superior to typical in-house teams or outsourced generalists. Centaur Labs recently launched an on-demand AI data labeling product that enables developers to collect expert labels in less than one hour. The company intends to use the new funding to expand operations and accelerate product development. Centaur Labs provides medical data labeling across images, videos, text, and audio by crowding a vetted network of tens of thousands of medical students and professionals. The company primarily captures labels via its gamified iOS app, DiagnosUs, which assesses labeler performance, rewards top contributors, and collects multiple opinions per case. Centaur combines multiple expert opinions into consensus labels, and the platform generates more than 1 million opinions each week from users in over 140 countries. Customers and collaborators include medical AI startups like Eko Health and research institutions such as Brigham and Women’s Hospital. The new capital will be used to expand the company’s global network of labelers, accelerate product development, and hire across the organization. Centaur participated in Y Combinator and emphasizes scaling its labeling infrastructure to support broad medical AI development.

Participated · Series A · Sep 2024

Treble provides a cloud-based sound simulation and synthetic audio data generation platform for applications across buildings, cars, tech products and immersive reality. The company was founded in 2020 by acoustic engineers Dr Finnur Pind and Jesper Pedersen and launched its platform in 2023 after close to a decade of research. Treble’s proprietary simulation technology has been benchmarked to be more than 100x faster than existing sound simulation solutions. Within a year of launch the platform is being used by some of the largest companies in tech, buildings, audio and automotive, including three of the five leading global tech giants and the world’s largest engineering consulting firms. The company plans to expand its team, enhance R&D, forge new enterprise clients and tap into new markets. Treble has raised €11m in a Series A to support those plans. Treble Technologies develops a cloud-based virtual sound platform that connects with major gaming engines and 3D modelling software to simulate sound and spatial audio in virtual environments. The company’s proprietary technology is aimed at industries including construction, automotive, consumer electronics, and the metaverse. Treble was founded in 2020 by acoustic engineers Dr. Finnur Pind and Jesper Pedersen and is based in Reykjavík, Iceland. Notable users include architecture studios Henning Larsen and Bjarke Ingels Group, as well as engineering firm COWI. The company plans to use the new funding to accelerate its digital sound capabilities and scale its virtual sound platform. The article does not disclose revenue or other operating metrics.