The disclosed venture funding announced across the wires on 28 July 2026 came to roughly $201 million. One check accounted for half of it. Strip out Coursera’s investment in LearnVector and what remains is eight rounds totalling around $101 million, with a median size under $7 million — a thin day by any measure, and a useful reminder that the headline aggregate in venture reporting is almost always one transaction wearing a crowd as a disguise.
The acquisitions were smaller still, and more revealing. A Canadian graphite developer agreed to be taken out at a 71% premium for an equity value of about C$11 million. A consumer brand changed hands for $5.5 million in cash. A retail computer vision company bought a competitor’s product line rather than the competitor. None of these are large transactions. All three describe a market where strategic value and market value have come badly unglued.
The Coursera Investment Is a Related-Party Transaction, and That Is the Story
Coursera announced a $100 million strategic equity investment in LearnVector, an AI-native learning company founded and led by Andrew Ng. The investment gives Coursera roughly a one-third ownership interest on a fully diluted basis, which implies a valuation for LearnVector in the neighbourhood of $300 million for a company incorporated in 2026 whose first product experiences are targeted for early 2027.
The detail most coverage will skip past is governance. Ng is Coursera’s chairman and co-founder. A public company investing nine figures into a private company controlled by its own chairman is a related-party transaction, and Coursera handled it accordingly: the board formed a special committee of directors determined to be independent and disinterested, that committee retained its own advisors, and it negotiated and unanimously approved the terms. That process is the correct one. It is also an admission that the transaction required insulation, which tells you how the board understood the optics.
Read commercially, the logic is defensible. Coursera brings accredited content, an enterprise and higher-education footprint reaching a very large learner base, and longitudinal data on learning behaviour. LearnVector brings agentic systems intended to plan personalised paths, adapt to individual learners, and persist through instruction, practice and assessment rather than terminating at an answer the way a chatbot does. The two parties describe the commercial collaborations as exploratory, not contracted — which is worth noting, because a $100 million equity investment with no binding commercial agreement attached is a bet on a person and a category, not on a pipeline.
Read defensively, it is an incumbent buying a call option on the technology most likely to make its existing catalogue business obsolete, at a moment when it is still digesting its Udemy combination. That is not a criticism. It is the correct trade if you believe the disruption is coming regardless. But it should be described as what it is rather than as a growth initiative, and the fact that the option was written by the chairman rather than sourced competitively is the part that deserves scrutiny.
Applied AI in Unglamorous Industries Took Most of the Remaining Capital
The single largest non-Coursera round went to COR, an Argentine company applying AI to profitability management at agencies and consultancies, which raised $30 million from FTV Capital. FTV is a growth investor rather than an early-stage firm, so this is a scaling round rather than a bet on product-market fit. The release went out simultaneously in English, Spanish and Portuguese, which suggests the round is being positioned as a marker for Latin American growth-stage technology as much as a company milestone.
Greyparrot raised $27 million in a Series B for computer vision applied to waste sorting, framed as infrastructure for the circular economy. The framing is doing promotional work, but the underlying business is sounder than the language: material recovery rates are a regulatory obligation across the European Union, and compliance spending survives budget cycles in a way that voluntary sustainability spending does not.
Curant.ai raised $3.1 million in a seed round led by Diagram to build AI for insurance. Diagram is an insurance-focused firm, so this is a domain investor backing a domain company — small check, narrow thesis, appropriate stage, and the kind of round that generates no follow-on coverage unless it works.
Frenos closed $1.52 million for AI-driven simulated penetration testing in operational technology environments. Operational technology security remains structurally undercapitalised relative to enterprise IT security, and simulation is the right product shape for environments where you cannot run live adversarial tests against production systems that control physical processes. The problem is hard and the raise is small, which is the usual configuration in industrial security.
Health, Hardware and the Very Small End
Epitel secured $26 million in a Series B to expand remote wireless EEG monitoring. This is device work rather than software, and it addresses a real gap: seizure monitoring still largely requires a hospital bed and a technician. The constraint on scaling is reimbursement rather than engineering, and nothing in the announcement addresses that.
Throne Science raised a $10 million Series A for a smart toilet sensor. The category invites mockery and mostly deserves it on presentation, but the mechanism is more interesting than the product description: passive health data collection that requires no user compliance whatsoever, which is precisely the failure mode that has limited every wearable ever shipped.
Olomon raised $2.6 million in an oversubscribed pre-seed to build what it calls a financial system of record for households. Positioning at that ambition level pre-product is aggressive, and household finance is a category littered with well-capitalised failures. The claimed differentiator is ledger completeness rather than budgeting interface, which is at least a different bet than the ones that did not work.
Liquid Interactive launched with $700,000 to let non-developers build visual-first AI experiences. It was the smallest disclosed round of the day, and the launch announcement and the funding announcement are the same document — usually a signal that the capital is the news because the product is not yet.
Eagle Point provided strategic financing to Microporous for a new manufacturing facility, with no amount disclosed. Battery separator manufacturing is capital-intensive and onshoring-adjacent, and structured credit funds are increasingly doing this work in place of banks. Separately, and outside the venture category entirely, Bending Spoons entered a €500 million term loan facility backed by SACE, the Italian export credit agency. Bending Spoons acquires mature software products and operates them for cash. Sovereign export credit supporting a serial software acquirer’s debt facility is an unusual structure, and arguably a policy story before it is a finance one.
Acquisitions: A 71% Premium on an Eleven Million Dollar Company
Lomiko Metals entered a definitive arrangement agreement to be acquired by Global Battery Materials for cash consideration of C$0.13 per share, by way of a court-approved plan of arrangement under British Columbia law. The consideration represents a 71% premium to the 20-day volume weighted average price on the TSX Venture Exchange through 27 July, and implies a fully diluted equity value of approximately C$11 million.
Both halves of that sentence matter. A 71% premium is a large number. C$11 million is a small one. Together they describe an asset — the La Loutre graphite project in Quebec — whose public market valuation had collapsed to a level where a strategic buyer could pay a premium that looks generous in percentage terms and negligible in absolute terms. Global Battery Materials is private, describes itself as building an ex-China supply chain from natural graphite through active anode materials, holds a prior producing mine in Canada and patented anode processing technology validated at pilot scale in South Korea, and states that it is backed by the Canadian government.
That is the graphite trade in miniature. Graphite is the battery input where Chinese export controls bite hardest, Western junior miners have been starved of capital for years, and consolidation into government-adjacent integrated platforms is the predictable result. The strategic case for the asset is stronger than it has ever been. The market price for it was C$11 million. Anyone writing about critical mineral security should sit with that gap rather than explain it away.
Elsewhere, Everseen acquired the Viztel portfolio from RadiusAI, extending its vision platform from checkout into full-store intelligence. This was an asset purchase rather than a company acquisition, which typically indicates the seller is refocusing or running short of runway. Retail computer vision is consolidating around a small number of platforms as standalone point solutions fail to reach the deployment scale that makes the unit economics work.
PSQ Holdings agreed to sell EveryLife for $5.5 million in cash, announced the same day it scheduled its second quarter earnings call. A small consumer brand sold for cash immediately before results is a balance sheet decision rather than a portfolio decision, and the price does not suggest a strong negotiating position.
Two Context Items Worth Reading Alongside
A separate release argued that corporate venture capital is entering a new growth phase as global companies turn to venture-capital-as-a-service — outsourcing the investing function rather than staffing it. That is the structural frame for the Coursera transaction and for a good deal of the strategic investment activity now being announced. When corporates outsource sourcing and diligence, deals that arrive through the chairman’s own network become relatively more likely, not less.
A second, pushed across ten languages by whoever paid for the distribution, found that nine in ten limited partners are more likely to commit to funds using leverage when disclosure is clear. The finding is not surprising. The survey is nonetheless citable for anything on NAV lending, subscription credit lines, and the disclosure practices that have quietly become the fund-raising bottleneck.
What the Day Actually Shows
Two hundred million dollars of announced venture funding, half of it in a single related-party transaction between a public company and its chairman. Eight other rounds with a median under $7 million, concentrated in applied AI for industries nobody writes about — agency profitability, waste sorting, insurance underwriting, industrial security. Three acquisitions, all small, one of them a strategically vital mineral asset changing hands for the price of a mid-sized seed round.
The capital is not absent. It is asymmetric, and it is going to companies solving narrow operational problems for buyers with regulatory obligations, while the assets that policy documents describe as critical trade at valuations that no policy document would predict.