Led · Seed · Jan 2026

Founded in November 2025 by Simone Conia and Edoardo Barba after years of research at La Sapienza University of Rome, Principled Intelligence builds technology that lets enterprises define their own operational principles and translate them into verifiable criteria for AI behavior. Its platform simulates realistic interactions, continuously monitors model outputs, and provides natural-language explanations for any violations, enabling transparent and controllable AI governance even for non-expert staff. The company previously created Minerva LLM, the first family of large language models trained from scratch on Italian data, which has logged more than 300,000 downloads across universities, companies and developers. Unlike existing governance tools, Principled Intelligence emphasizes easy customization via natural language and deep explainability, aiming to reduce reputational, regulatory and operational risks tied to generative AI deployment. Proceeds from the current financing will be used to expand the team and accelerate the rollout of its enterprise SaaS governance platform, targeting highly regulated industries. The founders argue that robust trust and oversight mechanisms will be critical for large-scale AI adoption, positioning the company to become a European standard-setter in AI control.

Led · Seed · Oct 2025

Supermemory is building a universal memory API designed to give AI applications long-term, highly personalized context. The system ingests virtually any data type—files, documents, chats, projects, emails, PDFs and live app streams—and constructs a knowledge graph that surfaces relevant insights on demand. Developers access the capability through a chatbot, notetaker and Chrome extension, with integrations into Google Drive, OneDrive and Notion to simplify data capture. Example use cases range from querying month-old journal entries to pulling relevant assets in a video editor, and current customers include Cluely, Montra, Scira, Composio’s Rube and Rets; a robotics company is also piloting the service for visual memory retention. Founder Dhravya Shah touts lower latency as Supermemory’s key competitive edge against peers like Letta, Mem0 and Memories.ai. The company recently secured $2.6 million in seed funding that will be used to enhance the product and expand developer adoption. Although specific revenue or usage metrics have not been disclosed, the capital provides runway for accelerating hiring and product development while refining an API-driven monetization model.

Participated · Seed · Aug 2025

Recornea is a clinical-stage Italian medtech company developing the GROSSO® Reshaper, a nitinol corneal implant designed to restore vision in patients with keratoconus. The GROSSO® Reshaper uses a predefined physiological curvature to remodel the cornea with predictable visual results and is built on a nitinol platform that the company says can be applied to other ocular conditions. Recornea completed preclinical work including biocompatibility testing and a six-month rabbit study showing the implant was well tolerated. A First-in-Human (FIH) study on 12 patients is underway at the Instituto de Microcirugía Ocular (IMO) in Barcelona, with a second FIH planned at Fondazione Policlinico Universitario Agostino Gemelli IRCCS in Rome; the first implanted patient showed excellent safety and promising early efficacy. The company is also exploring a nitinol-based filtering device for glaucoma as a second product in its pipeline. Funds recently raised will be used to complete the ongoing FIH studies and advance the pipeline toward larger clinical trials and commercialization. Recornea has developed GROSSO, the first nitinol corneal implant designed to help patients with advanced keratoconus recover vision. The implant can be inserted via a conservative-additive refractive surgery that does not require removal of corneal tissue. Recornea says its technology restores the physiological curvature of the cornea, a significant improvement over traditional intracorneal ring segments. Current standards cited in the article include corneal cross-linking for early disease and ICRS for advanced cases. The company aims to improve quality of life for keratoconus patients and integrate its solution into routine clinical practice. Recornea announced a €150,000 financing led by Scientifica Venture Capital to support these efforts.

Led · Equity · Jul 2024

CyLock has developed EVA, the Extended Vulnerability Assessment, which the company says surpasses traditional tests with targeted solutions to block hacker attacks. CyLock reports that customers and partners validated the solution from day one and have provided continuous feedback. The company intends to use new funding to continuously improve its product and to prepare for international expansion. Management emphasizes ongoing product improvement and growth driven by client and partner support. Financially, CyLock closed a €500k financing round that adds to €320k previously raised via the Cyber Xcelerator and other backers. Prior supporters included the Cyber Xcelerator (part of CDP Venture Capital SGR's National Accelerator Network), Zest Group, Innova Venture (the Regione Lazio VC fund), and several business angels.

Participated · Equity · Jul 2024

Fluid Wire Robotics is a research spin‑off accredited to Scuola Superiore Sant'Anna that develops robotic arms engineered for extreme environments including nuclear sites, space, underwater locations and explosion‑risk areas. The company was developed under the supervision of co‑founder Marco Fontana and is led by CEO and co‑founder Marco Bolignari. Its technology is positioned as a platform to provide sustainable, reliable robotic solutions for complex value chains such as nuclear energy, subsea exploration and space robotics. Fluid Wire aims first to develop and commercialize a robotic arm for the nuclear and space sectors, then to extend to underwater and explosion‑risk environments. The article reports a new €800k investment round that will be used to accelerate technology development and commercialization efforts. No revenue or user metrics are disclosed in the article.