The wire — 10 July 2026

Five rounds from 9 July that closed outside Thursday's brief. No single thread connects them — which is sometimes the accurate description of a news day. The most interesting tension sits between the smallest and third-smallest: a word processor that bets people will pay for AI that improves their writing without replacing it, and a workforce platform that bets enterprises will need an operating system for the AI agents they are about to hire. Two wagers on the same underlying question about AI and work; two opposite answers.

The largest round by a distance belongs to Cambridge's Alchemab Therapeutics, which has extended its Series A by €29.3 million (£25 million) with the British Business Bank as lead — the BBB's largest single life sciences investment on record. Existing investors SV Health Investors, Dementia Discovery Fund, RA Capital, DCVC Bio, Eli Lilly, Lightstone Ventures, and others all participated; the total Series A now stands at approximately £109 million (~€127.8 million). Alchemab's Resiliome platform mines antibody sequences from disease-resilient individuals — people who were exposed to a disease and didn't develop it, or recovered unusually fast — and uses AI to identify the protective patterns. The platform now holds 500 million sequences; the extension funds the path to one billion. The BBB's decision to lead is not primarily a financial signal, though the diligence it implies is real: it is a decision that the UK wants this platform retained domestically and is willing to put state capital behind that preference. The 18-month test is a Phase II entrant from the clinical pipeline.

Marker (London) raises €11.3 million ($13 million) in a Seed led by Index Ventures, with LocalGlobe and a set of specifically chosen angels: Steve Newman (Writely / Google Docs co-founder), Cal Henderson (Slack co-founder), and Thomas Wolf (Hugging Face). Marker is building a word processor. Not a text generator with a document interface — a word processor where AI assists ideation, drafting, and revision without producing the writing. The positioning is explicit: this is software for people who want to be better writers, not software that writes for them. The angel constellation is pointed: Newman knows what a disruptable document layer looks like, Henderson knows what happens to writing quality under collaboration pressure, and Wolf has invested in a product that pushes against the AI-generation output that Hugging Face's own infrastructure enables. CEO Jon Steinback was previously DeepMind's creative lead. The product question is whether an anti-generation positioning sustains a paid user base large enough to compete with incumbents that have enormous distribution advantages.

Aria (Paris) closes a €7 million Series A extension led by 115K, the venture arm of La Banque Postale. Separately, it has secured a €240 million debt facility for invoice purchasing — the balance-sheet fuel that powers its product. Aria embeds invoice financing into B2B marketplaces and platforms: a supplier invoices through the platform, Aria pays instantly (net of its fee), and Aria collects from the buyer on the original terms. The platform gets the financing feature without taking the financing risk. The €7M funds the AI underwriting infrastructure and team; the €240M is what Aria lends. 115K's entry is a diligence signal from an institution with institutional credit analysis capability. The 18-month test is deployment of the facility against a diversified enough client base that credit losses stay inside the underwriting assumptions.

Pixel-Flo (Sheffield) raises €6.1 million (£5.25 million) in a Seed led by Northern Gritstone, with SCVC, Parkwalk Northern Universities Venture Fund, and Germany's High-Tech Gründerfonds (HTGF). Pixel-Flo is a 2025 University of Sheffield spin-out developing a fluidic Continuous-Flow Mass Transfer process for microLED chip placement — an alternative to the pick-and-place robotics that degrade in yield and slow in throughput as chip sizes shrink toward consumer display dimensions. MicroLED's performance advantages over OLED (brightness, efficiency, longevity, no burn-in) are well understood; the manufacturing process has been the blocker. Pixel-Flo's Seed funds the move from lab validation to a commercial coating system demonstration — the step that turns a research claim into a supplier conversation with display manufacturers.

Munich's Sherpa raises €2 million ($2.2 million) in a Pre-Seed co-led by Seedcamp, DN Capital, Activant Capital, and Brighteye. Sherpa is building a platform to manage the full lifecycle of external work — from request to payment — covering contractors, consultants, and AI agents through a single interface. The thesis is that vendor management systems designed for human freelancers are a structural mismatch for AI agents, which have no timesheets to file and no invoices to submit. Whether that mismatch becomes acute fast enough to validate a new platform rather than a patch on existing VMS tools is the risk the Pre-Seed is taking. (Not to be confused with Sherpa.ai, the Basque enterprise AI platform ProYarn covered on 6 July.)

Alchemab Therapeutics (Cambridge, €29.3M / £25M Series A ext., British Business Bank). Marker (London, €11.3M / $13M Seed, Index Ventures). Aria (Paris, €7M Series A ext. + €240M debt facility, 115K). Pixel-Flo (Sheffield, €6.1M / £5.25M Seed, Northern Gritstone). Sherpa (Munich, €2M / $2.2M Pre-Seed, Seedcamp / DN Capital / Activant / Brighteye).