Participated · Equity · Jan 2019

Featurespace develops machine learning-based fraud and financial crime prevention technology for banks and payment service providers. The company is building privacy-preserving AI solutions that allow models to be trained on sensitive data without organisations sharing raw data. It will apply federated deep learning and privacy-enhancing techniques such as k-anonymity and local differential privacy to detect financial crime while protecting data privacy. Featurespace was awarded PETs Challenge Prize funding from UK and US government-backed programmes to build a prototype for privacy-preserving fraud detection. The PETs scheme was launched by Innovate UK and the US National Science Foundation and the project is backed by SWIFT. Featurespace has until 24 January to build its AI prototype, which if successful will be showcased at the US Summit for Democracy. Featurespace is a provider of Adaptive Behavioral Analytics and the ARIC platform, a real-time machine learning system that risk-scores events to prevent fraud and financial crime. ARIC uses explainable anomaly detection to identify suspicious activity and catch new fraud attacks in real time. The platform scores events in more than 180 countries and is deployed by over 30 major global financial institutions, including four of the five largest banks in the UK. Publicly announced customers include HSBC, TSYS, Worldpay, RBS NatWest Group, Contis, Danske Bank, ClearBank, Permanent TSB and Betfair. The company emphasizes its machine learning models’ ability to automatically adapt to shifts in consumer, business and criminal behavior. Featurespace intends to use new funding to support continued growth and to further develop its fraud and anti-money laundering solutions. Featurespace develops the ARIC™ real-time AI machine-learning platform that uses adaptive behavioral analytics and anomaly detection to risk-score transactions and other events. The platform is deployed in more than 180 countries and the company says it can reduce genuine transaction declines by as much as 70%. Customers include 17 banks across continental Europe, the U.K., the U.S. and Latin America, and several payment processors and merchant acquirers have embedded ARIC in their solutions. Featurespace was created out of Cambridge University’s Engineering Department and is led by CEO Martina King with CTO Dave Excell as a co-founder. The company has offices in Cambridge, London and Atlanta and says it has strengthened its senior management and operational infrastructure. Management plans to use new funding to accelerate international expansion and continue developing and distributing its software capabilities. Featurespace develops ARIC™, a real-time machine-learning platform that monitors individual behaviours and detects anomalies to identify risk and catch new fraud attacks. Developed out of the University of Cambridge and led by CEO Martina King, the company serves financial services and gaming organizations. Its products and services are deployed in over 180 countries and customers include TSYS, Playtech, Betfair, Vocalink Zapp, CashFlows, Camelot and William Hill. The company cites the fraud detection and prevention market was worth $14.23bn in September 2017 and is growing at roughly 20% per annum. Featurespace says it will use the funds to support international expansion and continued development of its software capabilities. Worldpay has also formed a commercial partnership to licence Featurespace’s behavioural analytics for merchant risk management and fraud prevention use cases. Featurespace provides a machine-learning behaviour analytics platform (ARIC) that models normal individual behaviour and detects anomalies in real time. The technology uses Bayesian statistics and is based on research at the University of Cambridge by the late Professor Bill Fitzgerald and Featurespace CTO David Excell. Led by CEO Martina King, the company serves customers across financial services and gaming, including Vocalink/Zapp, CallCredit, KPMG, Betfair, William Hill and Camelot. Featurespace focuses on real-time decisioning, monitoring each customer one at a time to deliver tailored fraud detection. The company intends to use recent funding to expand its operations in the UK and the US and to grow its presence in financial services. Financial metrics such as revenue or user counts were not disclosed in the article.

Participated · Series D · Sep 2017

SOPHiA GENETICS offers the SOPHiA DDM platform, which analyzes complex genomic and multimodal data to generate real‑time, actionable insights for hospitals, laboratories, and biopharma institutions. The platform is seeing global clinical adoption, with 463 core genomics customers and recent partnerships including Mayo Clinic going live for hereditary cancer and rare disease testing. In Q1 2024 the company reported $15.8 million in revenue, a 13% year‑over‑year increase, and operating losses of $18.8 million. Management reiterated full‑year 2024 guidance of 25%–30% revenue growth, an adjusted gross margin between 72.5% and 72.7%, and an adjusted operating loss of $45 million to $50 million. The company has arranged financing to support its growth initiatives and operations while continuing to expand access to data‑driven medicine. SOPHiA GENETICS states its mission is to use AI to deliver care for patients with cancer and rare disorders worldwide. SOPHiA GENETICS combines machine learning applied to medical imaging with genomic sequencing to generate a more holistic view of disease and improve care coordination. Its technology integrates multiple sources of medical data to surface novel insights about how diseases spread in the body. The Boston- and Lausanne, Switzerland-based company says its platform is used by more than 1,000 healthcare institutions and has analyzed 600,000 genomic profiles. The firm plans to use new funding to expand its footprint in the U.S. and Asian markets. Leadership changes — including adding former Agilent CFO Didier Hirsch to the board and creating an audit committee — suggest preparation for a potential public offering. The company emphasizes better patient care as its primary goal. Sophia Genetics develops SOPHiA AI, a universal platform that enables healthcare professionals to analyze complex genomic and radiomic data for oncology and hereditary disorders. Led by CEO and founder Dr. Jurgi Camblong, the company combines life sciences and medical expertise with advanced data computing and mathematical capabilities. Sophia Genetics is based in Lausanne, Switzerland, and Boston, MA. Its platform is utilized by more than 850 hospitals across 77 countries and has supported the diagnosis of over 300,000 patients. The company closed a $77M funding round and has now raised a total of $140M. It plans to use the funds to accelerate the reach of its data-driven medicine solutions and add talent to reinforce its presence in the US. Sophia Genetics operates a SaaS big-data analytics platform that aggregates clinician-fed genomic data and applies statistical inference and machine-learning algorithms to improve DNA-sequencing diagnostic accuracy. The platform ingests raw FASTQ data, has been validated on over 340 different sequencers, and has benchmarked samples from more than 10,000 patients and 500,000 unique variants. Hospitals connect and share annotated results so the AI improves as the user base grows — the company says it is adding about 10 new hospitals per month and currently works with over 350 hospitals. Accuracy has improved over time (from ~85% to ~98% in successive 10k-patient cohorts) with reported sensitivity between 99.9%–100% and specificity between 99%–100%. Sophia charges hospitals on a per-analysis, on-demand basis. With cheaper sequencing and richer datasets the company plans to expand analytic layers (for example combining imaging with molecular data) and to push deeper into oncology-case management and multi-modal predictive analyses. The new funding is intended to support global deployment and ramp commercial activity beyond Europe into Latin America, AsiaPac, Canada and the U.S. Sophia Genetics offers the Sophia DDM™ analytics platform, which hosts machine-learning algorithms and a large clinical genomics community to analyze genetic data for cancers and inherited conditions. The company supports more than 100 healthcare institutions across 20 countries, including Oxford University Hospital, Hospices Civils de Lyon, Vall d’Hebron Hospital and Cologne University Hospital. Led by CEO and co-founder Jurgi Camblong, the team comprises about 60 employees, two thirds of whom hold PhDs in genomics, bioinformatics, computational and data protection. Sophia leverages advanced analytics to help more accurately and quickly diagnose thousands of patients. The company plans to use new funding to continue market expansion and broaden its range of genomic tests. The firm has previously received backing from Mike Lynch, founder of Invoke Capital.

Led · Equity · Sep 2016

Luminance builds “legal-grade” specialist AI—described as a “Panel of Judges”—and offers products such as Lumi Go that can auto-negotiate draft agreements on customers’ behalf. The company uses a proprietary Legal Pre-trained Transformer (LPT) trained on over 150 million verified legal documents, many of them non-public, rather than general-purpose GPT models. Luminance emphasizes a mixed-model approach for accuracy and transparency and positions its platform across the entire contract lifecycle. It reports over 700 clients across more than 70 countries, including AMD, Hitachi, LG Chem, SiriusXM, Rolls‑Royce, and Lamborghini. The firm has significantly expanded in North America—tripling headcount there, opening offices in San Francisco, Dallas, and Toronto, and expanding its US HQ in New York. Financially, Luminance raised $75M in a Series C and, per CrunchBase, has now raised $138M to date; it was originally seed-funded by the late Dr Mike Lynch. Luminance builds a proprietary legal Large Language Model and a specialised legal co-pilot used across the contract lifecycle, from generation to post-execution analysis. Its flagship product, Luminance Corporate, and recent releases — a Legal-Grade™ chatbot for instant Q&A and automated redrafting, 'Self-Serve' for non-legal teams, and 'Auto Mark-Up' — target faster, standardized contract workflows. The company serves 700+ organisations in 70+ countries, including AMD, Hitachi, AB InBev, Panda Express, LG Chem, Koch Industries and Yokogawa. Luminance reports over 5x ARR growth in the past two years, generates more than one-third of its revenue in the U.S., and has seen a 225% increase in U.S. customers adopting its flagship product since January 2023. To support accelerated market capture, it has expanded its New York headquarters, added hires across North America (San Francisco and Canada), and opened a new Dallas office. Luminance develops an AI/ML platform for the legal industry that surfaces complex information in a visual dashboard to speed contract review. The product includes collaboration tools aimed at improving efficiency for M&A teams. The company launched in September 2016 and is led by CEO Emily Foges. It has deployed its technology across markets from South Africa to the Nordics and has attracted interest from law firms in Australia and the United States. Luminance has a workforce of 20 employees and was valued at over $20 million following the most recent funding. The company intends to use the funds to meet demand for its technology. Luminance is a London-based developer of a machine-learning platform designed to improve the legal due diligence process. The company leverages artificial intelligence and machine-learning capabilities developed at the University of Cambridge and anchored in Recursive Bayesian Estimation theory. Its document analytics software is built to handle forensic big data and add value to legal teams by freeing lawyers to focus on substantive work. Luminance has been collaborating with international law firm Slaughter and May to test and pilot the software for several months. Led by CEO Emily Foges, the company raised $3M and will use the funds to continue to develop the platform. Invoke Capital made the investment.

Participated · Equity · Mar 2015

Darktrace is a U.K.-based cyber defense firm whose proprietary technology sits on customer networks and uses machine learning to identify and stop malicious behavior. Its product is deployed on more than 7,000 networks and it counts customers including London Gatwick Airport, AIG, and the Science Museum Group. Since a $75 million Series D injection a year ago, the company says it has more than doubled deployments. Darktrace recently raised $50 million in a Series E that values the company at $1.65 billion. The company said the funding will drive further international expansion and support efforts to combat AI-based attacks. Headcount has increased by more than half and Darktrace has opened eight new offices, including Los Angeles and an expanded Singapore office, to meet growing demand. Darktrace builds an "Enterprise Immune System" appliance that sits on a company’s network, uses machine learning to identify malicious activity, alerts IT managers, and can take immediate action to stop or slow attacks. The company reports a total contract value of $200 million, 3,000 global customers, and 140% growth in the last year. Darktrace says it will use the new funding to expand business operations into more markets, including deeper expansion across Latin America and Asia Pacific. It has entered a strategic partnership with Hong Kong-based CITIC Telecom CPC to bring its technology to Asia Pacific (CITIC is a partner and customer, not an investor). Darktrace works with telcos such as BT and Telstra and is backed by investors including Autonomy founder Mike Lynch. Darktrace’s core product combines a box that sits on a company’s network with software that analyzes traffic using advanced mathematics and machine learning to distinguish normal from malicious activity. The system alerts IT managers to suspicious behavior and can take automated actions (for example throttling network speed) to slow or stop attacks. The company is developing technology to respond to human-written and emerging machine-learning-based attacks and emphasizes automation for rapid threats and ML for stealthy activity. Management says the backing will be used for further international expansion and research and development. The startup claims more than 1,000 customer deployments and reports revenue growth of more than 600% in the last year, translating to millions per month. Investors and early backers in the company include Invoke Capital and Hoxton Ventures, among others. Darktrace offers an enterprise immune system cyber‑defense platform developed out of the University of Cambridge that analyzes network activity and learns the behavior of every device, user and network to detect abnormal threat behaviors. The platform is positioned to protect enterprises and government bodies from advanced attacks and internal data compromises. The company serves customers across multiple verticals including energy and utilities, financial services, telecommunications, healthcare, manufacturing, retail and transportation. Darktrace plans to use the new funding to expand adoption—particularly in the United States and Asia Pacific—scale the business and hire technology, cyber analyst and commercial talent worldwide. The company was founded in 2013 and is based in Cambridge, UK and San Francisco, CA, with offices in Dallas, London, Milan, Melbourne, New York, Paris, Singapore, Toronto and Washington D.C. The Series B proceeds strengthen its go‑to‑market and product scaling efforts. Darktrace builds an 'Enterprise Immune System' based on new mathematics and machine learning developed at the University of Cambridge to learn normal network behaviour and detect subtle deviations in real time. Its product is delivered as a hardware box that sits on a network combined with software that analyses traffic, alerts IT managers to suspicious behaviour, and issues a regular 'Threat Intelligence Report'. The UK-based startup, founded in late 2013, says it addresses advanced insider and external cyber threats that rule-based systems miss. It claims rapid customer growth and counts Virgin Trains, BT and Drax among its customers. Darktrace is backed by Invoke Capital and other investors and reports rapid expansion plans, including moving into the Asia Pacific market. The company recently raised $18 million at an $80 million valuation.