The Race That Was Never Yours

Kairon Health

A dashboard can identify a care gap before breakfast. Closing it may still require a care manager, a field liaison, a scheduling system, a clinical record, an outreach list, and a patient who answers the phone.

That is the work Kairon Health just raised $5M to connect.

Flare Capital Partners led the Seed round, with Tau Ventures joining existing investors Lightbank, General Advance, and Pave Health Ventures. The capital will deepen Kairon's five workflow clouds, care operations, patient outreach, and go-to-market.

Co-founders Nick Bartz, CEO, and Evan Gogel, President, brought operating experience to the problem. Nick Bartz spent 8 years in business intelligence and field operations at Aledade. Evan Gogel was a co-founding partner at Redesign Health and the first CEO of Vault Health.

They are making a precise bet: value-based care does not suffer from a shortage of dashboards. It suffers in the distance between seeing a problem and completing the work that changes it.

Kairon links claims, clinical records, labs, pharmacy data, meeting transcripts, field notes, and CRM logs into a patient-to-practice model. Practice Intelligence prepares teams for the conversation. Care Operations coordinates worklists. Patient Activation books appointments. Portfolio Management supports network decisions. Impact Analytics measures whether an intervention reached a billed outcome.

Kairon says the system is live across MSSP, ACO REACH, Medicare Advantage, Medicaid, and commercial contracts covering more than 1M attributed lives in 30+ states. Two unnamed deployments each manage about 350,000 lives. Those are company-reported adoption figures, not an independent outcomes audit.

The timing gives the round more weight than the amount suggests. CMS says 14.3M Medicare beneficiaries were receiving ACO-coordinated care in 2026. ACO REACH ends in December, and the 10-year LEAD model begins in January. Risk-bearing organizations will face new benchmarks with the same finite staff hours.

Kairon now has to prove that one connected system can reduce those handoffs without becoming another place where work waits. Data must stay reliable, priorities must fit clinical capacity, outreach must become completed care, and measurement must survive scrutiny.

The care gap was never the only gap. The missing layer sits between knowing what should happen and giving a stretched organization enough coordinated capacity to make it happen.

Deeper Analysis 👇 https://devcuration.com/articles/kairon-health-raises-5m-value-based-care-execution

Nix Biosensors

A sweat patch has one job and a short life. The dataset behind it can travel much farther than the adhesive.

That is the shape of Nix Biosensors' $10M Series A. The Boston health-tech company has spent years measuring fluid and electrolyte losses in the field, then turning those readings into personalized hydration guidance for athletes, military personnel, workers, coaches, and performance teams. The round brings company-reported total funding to $17M.

Founder and CEO Meridith Cass reached hydration through a hard lesson. Nix traces its origin to Cass experiencing hyponatremia during the Boston Marathon and finding little personal data for a body under stress.

Nix built the collection system first. Its bicep-worn sensor takes continuous sweat readings. Nix Solo guides individuals; Nix Pro lets coaches, nutritionists, and supervisors monitor groups. Its partner record spans collegiate programs, U.S. Speedskating, professional cycling, soccer, motorsports, military settings, and labor environments.

The new capital is aimed at the handoff from device to intelligence. Nix Health is being developed as a hydration algorithm that can be licensed into existing fitness and consumer-health platforms without requiring a Nix sensor. The patch created the observations. The commercial bet is that the learned model can meet customers wherever their health data already lives.

Shorewind Capital, Great Oaks Venture Capital, Oregon Sports Angels, and White Road Investments joined the round, alongside Anne Wojcicki, Len Blavatnik, and owners of 5 professional sports franchises. The announcement does not name a lead investor, valuation, or detailed terms.

That roster makes sense for a company sitting between sports performance, occupational safety, consumer health, and software distribution. The difficult asset is a field dataset built across different bodies, conditions, activities, and organizations, then translated into guidance someone can use while the body is still losing fluid.

Nix also plans to explore other sweat biomarkers. Those are development directions, not finished products. The nearer test is whether Nix Health can turn years of sensor deployment into a trusted licensing layer that platforms will put in front of millions of users.

The patch may remain the sharpest way to collect the signal. This Series A gives Nix room to learn how far the intelligence can move after the workout ends, the sensor comes off, and the data has to earn a place inside somebody else's platform.

Deeper Analysis 👇 https://devcuration.com/articles/nix-biosensors-raises-10m-hydration-intelligence

VerifAIX, Inc.

Somebody eventually has to put a name next to the word “correct.”

That burden gets heavier when AI can generate specifications, RTL, assertions, and testbenches faster than a verification team can explain why any of them should be trusted. A plausible answer is useful in a chat window. A plausible chip can become an extraordinarily expensive object lesson after tape-out.

Madhulima Tewari, Kenneth Roe, and Avner Landver built VerifAIX around that gap. The company has raised a $5M Seed round co-led by Endiya Partners and Bluehill.VC, its first institutional financing.

VerifAIX is not asking another model to grade the model’s homework and call the stack intelligent. Its platform begins with specifications, RTL, and existing verification assets, then builds what the company calls a Formal Brain: a mathematically grounded representation of what the design is supposed to do.

That shared model supports verification planning, testbench and assertion generation, formal analysis, simulation, coverage, debugging, and closure. Every downstream artifact is meant to trace back to design intent, while engineers keep approval authority and the platform works through EDA tools already inside the workflow.

The company says its platform is already being used in pilots and deployments with multiple semiconductor companies on control-intensive and protocol-heavy designs. Those customers are unnamed. Revenue, deployment counts, and audited productivity remain undisclosed.

The capital will deepen the Formal Brain, scale automated abstraction and decomposition to larger designs, strengthen the connection between formal verification and simulation, expand customer deployments, and grow engineering teams across the US, India, and Israel.

Endiya Partners and Bluehill.VC are underwriting more than a tool launch. They are betting that AI-assisted chip development creates a second market beside generation: independent evidence that the growing pile of machine-produced work still matches the intent humans approved.

The founders now have to move that evidence from complex blocks and IP into larger subsystems, integrate it with workflows chip teams already trust, and make rigor feel like acceleration instead of another layer of process.

If that handoff works, the most valuable AI in semiconductor design may not be the system that writes the next artifact. It may be the one that helps an engineer sign the design off without crossing their fingers.

Deeper Analysis 👇 https://devcuration.com/articles/verifaix-raises-5m-chip-verification-trust-layer

Payment Nerds

Shawn Silver spent more than a decade in merchant services before founding Payment Nerds. The $2M credit facility now backing the company is aimed at the part of the business a landing page cannot scale: sales judgment, integrations, partnerships, infrastructure, and people.

Payment Nerds secured the non-dilutive facility from Espresso Capital on September 14. The capital is expected to support sales and marketing, technology and integration capabilities, strategic partnerships, merchant services, and hiring while preserving the company's existing equity structure.

The rate, maturity, security package, covenants, and draw schedule were not disclosed. Neither were revenue, merchant count, transaction volume, or valuation. This is financing capacity, not evidence that every dollar has been drawn or every growth target has been met.

Merchant services is a business where a successful transaction is only the visible moment. Merchants also need underwriting, predictable deposits, rational reserves, fraud and chargeback controls, reconciliation, accounting connections, and somebody who answers when those pieces disagree.

Payment Nerds sells across that stack: card, debit, wallet, and ACH payments, plus gateways, POS systems, invoicing, fraud controls, treasury tools, QuickBooks connectivity, and support for specialized merchants.

Shawn Silver previously founded PaymentCloud. In April 2025, Electronic Merchant Systems, now Kurv, made an undisclosed strategic investment in Payment Nerds and entered a distribution agreement with the company. The Espresso facility adds debt capacity to an operating network that already includes a processing partner.

The people closest to the capital's first job are already in place. Payment Nerds appointed Jacob Martin as VP of Sales in August and promoted Trae Holthouse to Sales Manager. Martin is building a repeatable sales and partner strategy; Holthouse is developing the team and tightening prospecting, closing, and retention.

Debt preserves ownership, but it also puts a calendar on execution. Sales hires need productive pipelines. Integrations need active merchants. Partnerships need dependable distribution. Support has to survive added volume without becoming the reason a merchant starts shopping again.

That is the handoff inside this $2M facility. Payment Nerds is financing the machinery around the payment, then asking merchants and partners to trust that machinery when money, risk, and customer experience meet in the same transaction.

Deeper Analysis 👇 https://devcuration.com/articles/payment-nerds-secures-2m-credit-facility-for-growth

Ayble Health

Chronic digestive care has a strange org chart: the patient becomes the project manager.

Diagnosis, medication, nutrition, and behavioral health often sit with different people. The person managing the symptoms gets the extra job of carrying context between them.

Sam Jactel built Ayble Health after ulcerative colitis made that fragmentation personal. The company has now closed an oversubscribed $16M Series A led by Neon, with Unum Ventures, Upfront Ventures, M13, Cleveland Clinic Ventures, DigiTx, Accomplice, and strategic individuals participating. Ayble reports more than $27M raised in total.

Ayble plans to deepen AI-supported personalization for engagement, enrollment, and care delivery, then extend its multidisciplinary clinic from GI care into autoimmune conditions.

Ayble brings physicians, nurse practitioners, dietitians, behavioral-health specialists, pharmacists, coaches, and care coordinators into one virtual model. The AI matters when it helps that team direct care without turning another dashboard into the patient's homework.

Neon's founding and managing partner Kimmy Scotti pointed to the sequence behind the investment: Ayble spent years building clinical evidence before scaling the business.

Digital health has produced plenty of distribution stories that asked evidence to catch up later. Ayble arrives with 16 peer-reviewed publications, a place in Peterson Health Technology Institute's 2026 virtual-GI assessment, and a 2025 prospective study in which 86% of 202 participants achieved a clinically meaningful reduction in IBS symptom severity.

The evidence has boundaries. The 2025 study was uncontrolled and recommended randomized trials. PHTI found promise across clinician-led virtual GI care while calling for stronger comparative, durability, adherence, and diverse-population research.

Ayble says it reaches more than 80M covered lives and reports a 47% average symptom improvement within four weeks, sustained for at least a year, plus a minimum 3:1 ROI across its book of business. Those are company-reported results. Covered lives describe access, not active use.

That is where this Series A earns its next chapter. Ayble must turn eligibility into enrollment, enrollment into coordinated treatment, and GI evidence into an autoimmune-care model that holds up across more conditions and handoffs.

The patient has already spent enough time running the meeting. This capital gives Ayble a larger obligation to make the care system own the agenda.

Deeper Analysis 👇

https://devcuration.com/articles/ayble-health-raises-16m-precision-gi-care

Sling Therapeutics

Five infusions can look efficient in a biotech deck. They are still 5 appointments for a patient whose eyes may already hurt, bulge, or see double.

Thyroid eye disease now has 2 approved IGF-1R biologics in the U.S. Both are delivered by infusion. Sling Therapeutics is financing the oral argument.

The Ann Arbor biotech closed a $123M Series C led by Forbion, with existing investor TPG Life Sciences Innovations and new investor Sectoral Asset Management participating. The capital arrived with a clinical handoff already in motion: Sling dosed the first patients in ORBIT, its Phase 3 trial, on the same day it announced the round.

Linsitinib is an oral small-molecule inhibitor of IGF-1R, the same target behind the approved biologics. Sling licensed the drug from Astellas and launched in 2022 with a $35M Series A led by TPG's The Rise Fund. Founder Gary D. Hammer and CEO Ryan Zeidan have moved a former oncology asset into a disease where the biology is validated and the treatment routine remains negotiable.

The company-reported LIDS study enrolled 90 patients across 35 sites in 5 countries. Sling said the 150 mg twice-daily arm produced a 52% proptosis response at Week 24 and met the primary endpoint with p=0.01. It also reported a favorable profile across hearing, glycemic, and menstrual-cycle changes.

Useful signal, serious qualification. Those are sponsor-reported topline results, and linsitinib is not approved.

ORBIT is where the oral thesis meets the pivotal standard. The trial plans to enroll about 130 adults with moderate to severe active TED, receiving linsitinib or placebo twice daily for 24 weeks. Proptosis response is the primary endpoint, followed by an open-label extension for eligible patients.

The market has moved. Amgen's Tepezza established IGF-1R commercially. Viridian's Lumvoa joined it in June with a 5-infusion course. Convenience is becoming part of the competitive design.

The $123M is paying for confirmation, differentiation, manufacturing, regulatory execution, and commercial preparation. Terms and valuation are undisclosed.

Forbion, TPG, and Sectoral are backing a narrow proposition with a large operating burden behind it: prove that a pill can deliver enough efficacy and safety to change how specialists and patients navigate TED treatment.

ORBIT now moves that proposition into 130 patient schedules, 24 weeks of dosing, blinded measurements, and the regulatory file that will decide whether oral convenience earns a place beside clinical benefit.

Deeper Analysis 👇 https://devcuration.com/articles/sling-therapeutics-raises-123m-oral-ted-therapy

Noetive

A factory can know where every pallet, machine, worker, and order sits, then still depend on one veteran operator to decide what happens when the plan breaks.

Noetive has raised a $41M Seed round to turn more of that judgment into software.

Eclipse led the financing, with Craft Ventures, The Westly Group, Swish, Factory, Incite, Gigascale, Operator, and Liquid 2 participating. CEO Amir Frenkel and CPO Dan Barak founded Noetive this year to build an "intelligence of record" for the physical economy.

Industrial businesses already have ERP, warehouse, transportation, sensor, spreadsheet, and paper records. Trouble arrives when delayed material changes production, labor, inventory, delivery, and the customer promise at once.

Noetive's agents sit on top of existing tools while multimodal sensing pods read the physical environment. The system is meant to learn the operation, absorb working knowledge, take on end-to-end tasks, and improve the next decision.

The first wedges are specific: dynamic production planning in manufacturing and manifest consolidation in logistics.

Steuben Foods CEO Menachem Katz says planning that once happened monthly and took a week now happens daily and takes minutes. The claim is company-reported, not an audited study, and Noetive has not disclosed customer count, pricing, revenue, retention, savings, or error rates.

Eclipse did more than lead the round. The firm says it built Noetive with Frenkel after he joined as its first chief AI officer, helping shape the thesis, assemble the team, and connect the company with industrial problems.

That model puts an uncomfortable requirement inside the product. Industrial intelligence cannot stay impressive in a lab. It has to understand the local exception, work through existing systems, earn operator trust, and make a better call when the machine stops or the order changes.

Noetive will use the capital for frontier research, field deployments, and hiring AI researchers and full-stack engineers. Valuation, terms, ownership, prior funding, and a current CTO were not disclosed.

The $41M gives the team room to build a repeatable operating layer. The proof will be whether knowledge learned beside one crew becomes useful across the next plant without turning every deployment into a custom project.

Production planning and manifest consolidation are the opening assignments. The larger company gets built one decision after the floor changes.

Deeper Analysis 👇 https://devcuration.com/articles/noetive-raises-41m-physical-operations-ai

Rune

A data center usually begins with permits, a grid request, concrete, transformers, and years of expensive patience.

Rune wants to begin with a forklift.

William Layden and Varun Palivela have raised a $40M Series A led by Spark Capital for RELIC, a modular data center that sits inside a solar plant and connects to the direct-current power those panels already produce. Union Square Ventures, Lowercarbon Capital, Activate Capital, Committed Capital, Timeless Partners, and Logos Fund joined the round, bringing reported total funding to $53.5M.

The interesting accounting happens before the capital reaches Rune. Solar plants produce electricity that can be clipped by inverters, curtailed when the grid cannot take it, or stranded behind transmission limits. Meanwhile, AI companies wait for conventional data-center capacity that needs more generation, grid work, transformers, and construction.

RELIC connects behind the inverter on the solar plant's DC bus. Servers also run on DC power, so Rune can avoid part of the conversion and transmission stack, place its 8-by-8-foot modules between panel rows, and sell the resulting compute. The company owns and operates the equipment and is initially focused on inference workloads.

That turns 2 balance-sheet problems into the same operating question. A renewable owner wants another buyer for power that would otherwise earn nothing. A compute customer wants GPUs before a multi-year infrastructure schedule becomes obsolete.

Rune says a module can be installed in 60 minutes and customers can reach compute within 6 weeks of contract. It reports about 1MW of current capacity across Texas and California, including equipment at a 200MW Texas solar facility, with 80MW+ of contracted power.

Those are Rune's numbers, not independently audited outcomes. Revenue, pricing, margins, valuation, ownership, customer concentration, and the Texas solar operator remain undisclosed. Rune still has to prove that intermittent energy, inference workloads, hardware reliability, and site economics can live inside the same contract at scale.

The $40M gives Rune room to manufacture more systems, add power and compute partners, hire across infrastructure, and expand toward wind. The deeper bet belongs to William Layden and Varun Palivela: data-center development can behave more like manufacturing when the machine travels to available power.

Every new RELIC will test how much valuable infrastructure has been waiting behind a solar inverter for somebody to give it another job.

Deeper Analysis 👇 https://devcuration.com/articles/rune-raises-40m-ai-compute-solar-farms

Footprint

Eli Wachs and Alex Grinman are asking banks to let software inherit something usually kept in people’s heads: the reason a difficult compliance decision was made.

Footprint has raised a $25M Series B led by QED Investors to expand Percy, its AI operating system for financial-crime compliance, and Trust Fabric, the governed institutional memory beneath it.

The financing matters because financial-crime work is not short on alerts. It is short on usable context.

An analyst clears a watchlist match, escalates an adverse-media result or approves a customer after enhanced due diligence. The reasoning may be sound, but too often it remains trapped in a case note or the memory of the person who made the call.

That makes future reviews more expensive and consistency harder to prove when an auditor asks why two similar cases produced different outcomes.

Footprint’s thesis is that AI should not simply produce another answer. It should preserve the evidence, policy and prior decisions that shaped the answer, then let the institution reuse that context under human control.

Percy’s agents cover KYC and KYB, sanctions, transaction monitoring, enhanced due diligence and adverse media. They collect evidence and propose an action. A person approves it, with citations, timestamps and a task trail available for review.

Trust Fabric is the deeper bet. It turns the reason behind a compliance decision into governed memory that can inform the next investigation.

Remembering why the organization acted is a different product from automating a checklist.

Footprint says its software has reduced some watchlist reviews from 30 minutes to less than one minute and some enhanced-due-diligence reviews from three hours to under 15 minutes while finding 35% more evidence. Those figures are company-reported, not independently audited, but they clarify the pitch: speed that still leaves a defensible record.

The company says customers include regulated banks, Bilt, Nuvei and MoonPay. It plans to double its engineering and sales organizations and open a San Francisco office.

There is a tension worth watching. The more judgment a platform absorbs, the more institutions must govern which precedent is trusted, how conflicting decisions are resolved and where human accountability begins. Institutional memory can compound good judgment. It can also compound an old mistake.

Footprint is betting that auditability is not a constraint on compliance AI. It may make adoption possible.

Deeper Analysis 👇 https://devcuration.com/articles/footprint-raises-25m-auditable-financial-crime-ai

Bain Capital Ventures

Thirty to 40 companies will decide whether “post-AGI” becomes an investment category or an unusually ambitious fund memo.

Bain Capital Ventures has closed Fund XI above target with $1.6B in total capital. The September 16 announcement gives BCV its 11th core vehicle and a larger pool than the roughly $1.4B Fund X it raised in 2023.

The phrase doing the heavy lifting is “post-AGI.” BCV is underwriting a world where artificial intelligence moves from impressive capability to economic infrastructure. The firm is looking across infrastructure, applied AI and services, physical AI, health, and security for the companies that make that transition usable in the real world.

That scope could become a shopping list. BCV has imposed a more useful constraint: Kevin Zhang told TechCrunch the firm expects to back 30 to 40 companies, primarily from Seed through Series B. BCV also says partners often work in pairs or trios on an investment so each team receives enough attention.

The company count matters because $1.6B can make capital look abundant. Partner judgment is not. Neither are founders capable of turning cheaper intelligence into dependable products, physical systems, healthcare workflows, security controls, and infrastructure customers can trust.

BCV says more than 82% of Fund X dollars went into Pre-Seed through Series B rounds. That is an allocation metric, not a return figure, but it makes the early-stage claim measurable.

Fund XI also carries the larger Bain Capital platform behind it. BCV points to access across credit, real estate, insurance, private equity, and operating relationships as an advantage for founders whose ambitions require more than another equity check. Bain Capital's approximately $225B in assets is platform context, though, not extra money inside Fund XI.

The LP disclosure shows alignment without naming institutions. Bain partners, employees, and related entities are among the fund's largest investor groups, alongside pensions, endowments, and foundations. Fees, carry, check sizes, deployment pace, and performance expectations were not disclosed.

Slater Stich, Christina Melas-Kyriazi, and Kevin Zhang attached their names to the official thesis. The capital now turns that thesis into a sequence of choices: which layer of AI gets funded, which founders receive concentrated partner time, and which companies can carry technical ambition into markets where reliability still gets the final vote.

Deeper Analysis 👇 https://devcuration.com/articles/bain-capital-ventures-closes-1-6b-fund-xi-ai

The Carlyle Group

Infrastructure projects and standard loan documents live on different clocks.

A data center may need power before the grid can deliver it. A transport asset can spend years moving through permits, construction, contracts and demand risk. The financing has to understand the asset before the asset can fit the financing.

The Carlyle Group has closed Carlyle Infrastructure Credit Fund II with approximately $2.3B in total capital commitments. CICF II exceeded its $2B target and is more than three times larger than its predecessor, according to the firm.

The size gets the headline. The more useful detail is that Erik Savi's team had already assembled six investments across North America and Europe with about $500M committed when Carlyle announced the final close.

Carlyle is building around directly originated, privately negotiated financing for infrastructure businesses and assets, seeking below-investment-grade opportunities across energy transition, digital infrastructure, transportation and logistics, low-carbon power, water and waste treatment.

That range creates the opportunity and the underwriting problem. Essential infrastructure can carry durable demand, but a data-center financing does not behave like a water asset, and a transport business does not inherit the same risks as low-carbon generation. Flexible credit earns the word flexible only when the structure changes with the asset and the discipline survives the change.

Mark Jenkins framed the close as evidence of investor demand for differentiated private-credit strategies. Carlyle now reports approximately $8.7B across Infrastructure Credit and $211B across Global Credit as of June 30, 2026.

The outside market is supplying plenty of work. The IEA expects global energy investment to reach $3.4T in 2026 and says annual grid investment must rise about 50% by 2030 from roughly $400B today. More than 2,500 GW of renewable, storage and large-load projects remain stuck in grid queues.

This is where private credit specialization gets tested in documents, covenants and monitoring, not in the size printed on a fundraising announcement.

CICF II gives Carlyle more room to finance the gap between infrastructure ambition and the slower machinery that makes an asset bankable. The next six investments will tell us more than the first six about whether scale can keep the underwriting close to the steel, permits, contracts and cash flows underneath it.

Deeper Analysis 👇 https://devcuration.com/articles/carlyle-closes-2-3b-infrastructure-credit-fund-ii

Artemis

The most important machine in a strategic industry is often the one nobody outside the factory can name.

A radar system depends on RF components that survive failure-intolerant conditions. A new therapy may depend on a freeze-dryer preserving its stability. A semiconductor tool is only as trustworthy as the optical engineering and metrology proving that it works.

Artemis Capital Partners has closed Artemis Capital Partners IV at $254M. The fund reached final close on September 1 after entering the market with a $250M offering target in its SEC filing.

James Ward, CEO and Partner, and Euan Milne, CIO and Partner, are putting that capital behind the industrial technology layer serving aerospace, defense, life sciences and semiconductors.

That layer is technical, fragmented and easy to overlook until one component misses tolerance, one instrument loses accuracy or one supplier cannot meet a qualification requirement.

Artemis says Fund IV had acquired three platforms by the close. The public deal record shows the thesis in motion: Diamond Antenna and Microwave makes RF rotary joints and subsystems for radar, satellite communications, air traffic control and electronic warfare. Millrock Technology makes freeze-drying instruments for pharma, biotech and diagnostics. Optikos designs, builds, tests and validates optical systems for defense, space, life sciences and semiconductors.

The final-close release does not name those three businesses as Fund IV assets, so the public record should not pretend it does. Their timing and capabilities still make the operating idea clear.

Specialization is not one playbook printed three times. RF reliability, lyophilization process control and optical metrology ask different questions of product development, sales, manufacturing and customer trust.

Artemis is betting that a private equity team can learn those languages deeply enough to help small industrial companies expand without sanding away the engineering discipline that made customers rely on them.

The $254M gives the firm more room to buy. It also raises the number of ownership decisions that have to be right across hiring, equipment, add-on acquisitions, quality systems and market expansion.

The close establishes the mandate. The work now sits inside businesses whose products may remain invisible even while the aircraft, therapies, chips and instruments around them become impossible to ignore.

Deeper Analysis 👇 https://devcuration.com/articles/artemis-closes-254m-industrial-tech-fund-iv

Brighteye Ventures

An investment category usually outgrows its label before the institutions around it agree on a new one.

Brighteye Ventures built its reputation in European edtech. Its $72M first close for Fund III is a decision to follow learning beyond the classroom, into the systems where people apply judgment, navigate work, find opportunity, and adapt while AI changes the job underneath them.

The first close brings Brighteye's company-reported AUM to $245M. New LPs include Lumina Foundation, Zanichelli Venture, and PI Impact, joining returning investors European Investment Fund, Jacobs Foundation, and several European family offices. The fund target, final-close timing, and individual commitments were not disclosed.

Founding Partners Ben Wirz and Alex Spiro Latsis are calling the expanded thesis the HumanOS: technology that helps people learn, work, and adapt. The phrase earns its keep when you look at the portfolio evidence. Fund III has already backed imagi for AI fluency in schools, NEX Health Intelligence for hospital infection-control teams, and Gyver for the electrician labor market supporting Europe's energy infrastructure.

Those companies do not sit inside one tidy software category. They sit around the point where intelligence has to become human capability before it creates economic value.

That distinction matters now. Brighteye's own research says European Learning & Work funding rose from €710M in 2024 to €1.6B in 2025, then reached €1.4B in H1 2026. The capital is returning, but it is also concentrating around companies tied to regulated workflows, institutional buyers, labor-market transactions, and measurable operating outcomes.

Fund III gives Brighteye room to make up to 35 investments. Eight have already been made. The firm also promoted David Guérin to Partner, Isabella Vahdati to Principal, and Rhys Spence to Head of Platform & Research, adding operating depth while the mandate gets wider.

That is the live challenge inside this close. Sector specialization creates value because the investor sees patterns a generalist misses. Category expansion creates value only if the same judgment still connects a school, a hospital, a productivity layer, and a skilled-labor marketplace without turning "future of work" into a warehouse for unrelated ideas.

Brighteye now has $72M of first-close evidence behind that wider frame. The next proof will arrive company by company, wherever human judgment becomes more valuable because the software around it got smarter.

Deeper Analysis 👇 https://devcuration.com/articles/brighteye-ventures-72m-fund-iii-first-close

Ridgeline

A customer writing an equity check into the system that runs its business is a different kind of product review.

That is the sharpest detail in Ridgeline's $250M Series E. Founder and Chairman Dave Duffield led the invitation-only round at a $1.425B valuation. Motley Fool Ventures, associates of Smead Capital Management, and Patrick O'Shaughnessy also participated as customers or affiliates.

Ridgeline is asking asset and wealth managers to replace a familiar mess: trading, accounting, compliance and reconciliation in separate systems, followed by reporting that has to explain why the numbers disagree.

The company puts those functions on one real-time data model. Ridgeline Intelligence adds assistants, workflows and agents that can prepare client meetings, investigate breaks and support pre- and post-trade compliance.

That architecture matters because an AI agent in finance needs more than a clever model. It needs permissioned context, an audit trail, current books and records, and a human who can see why the system acted. FINRA's 2026 oversight report makes the operating burden plain: governance, testing, monitoring, documentation and human review still follow the work.

CEO Dave Blair is selling capacity into an industry where growth has often dragged more people, systems and cost behind it. Ridgeline says customers consolidate an average of 6 to 9 legacy systems and that more than $750B in AUM/AUA is committed to the platform. Those figures are company-reported, and committed assets are not the same as independently audited production economics.

The investor list gives the numbers more texture. Motley Fool Ventures said its conviction benefited from hearing directly from colleagues who use Ridgeline. Cole Smead described seeing the platform from both the customer and personal-investor sides. Professional investors are used to judging software from the outside. Here, part of the diligence happened while the product was doing the work.

Ridgeline will use the capital to extend AI capabilities, expand managed services, build customers in Canada and Europe, and continue product development. It has not disclosed revenue, customer count, round terms, prior funding, ownership changes or independently audited productivity results.

The $250M buys room to grow. Customer capital raises the obligation: the system now has to carry more firms, more jurisdictions and more agentic work without losing the controls that made those customers trust it in the first place.

Deeper Analysis 👇 https://devcuration.com/articles/ridgeline-raises-250m-investment-management-ai

Impulse Space

The order book reached 2028 before Helios reached orbit.

Impulse Space has sold out Caravan 2 and 3, Helios rideshare missions planned for 2028. Helios is targeting its first flight in 2027. The distance between demand and delivery is why this $308M Series D extension matters.

The new capital takes the round to $808M after the $500M tranche announced in June. Returning investors 137 Ventures, BANNER VC, DFJ Growth, Linse Capital, Lux Capital, and Valor Equity Partners participated. Impulse Space did not name a lead for the extension or disclose terms. Reuters, citing a source close to the company, reported a $5.4B valuation.

Tom Mueller, founder, CEO & CTO, built Impulse around the mobility problem after launch. Mira hosts, deploys, and maneuvers payloads. Helios pushes larger payloads toward high-energy orbits. Electra adds efficient electric propulsion alongside the chemical fleet.

The product map is ambitious. The operating problem is brutally physical.

Reuters reported more than 600 employees and over 180 open roles. President & COO Eric Romo said the company raised the extension so it could hire ahead of customer plans, even if demand arrives more slowly than expected. Payload reported roughly six Mira vehicles in the factory while Impulse prepares to increase Helios production.

Government demand is adding pressure. Space Systems Command placed Impulse in NSSL Phase 3 Lane 1 with a $5M on-ramp task order, making it the program's first upper-stage prime. A separate $28M extension supports two more VICTUS SALO missions using Mira. Those awards are evidence of procurement interest, not completed mission performance.

The company also hired Adam Townsend as its first CFO. That may sound less cinematic than a rocket engine, but a company with at least $1.333B in publicly disclosed rounds needs a financial operating system as disciplined as its propulsion program. Townsend previously helped take VIZIO public and through its sale to Walmart.

The $308M is buying room to build before every customer schedule becomes urgent at once. It does not remove the work: Helios still has to fly, production has to hold, and backlog has to become revenue without hardware complexity eating the margin.

Impulse Space now has capital, contracts, people, and a sold-out future manifest pulling on the same production line. The next few years will be measured in qualified engines, finished spacecraft, and missions delivered into a calendar customers filled early.

Deeper Analysis 👇 https://devcuration.com/articles/impulse-space-extends-series-d-by-308m-for-orbital-scale

Comp AI

One million users taught Lewis Carhart, Claudio Fuentes, and Mariano Fuentes when to stop building.

Their previous startup, LeapAI, crossed that mark and still failed to find a use case sticky enough to justify more investment. It left them with a less glamorous memory: getting through SOC 2 by hand while trying to build a company.

That problem became Comp AI. Now it has become a $34M Series A co-led by Roo Capital and Grand Ventures.

Lewis Carhart leads the company as CEO, Claudio Fuentes as COO, and Mariano Fuentes as CTO. Comp AI says it serves more than 1,000 customers and grew ARR 15x year over year. Those company-reported figures explain why the firms are financing more than audit preparation.

Comp AI's agents gather evidence, generate policies and risk records, monitor controls, and support vendor assessments. The company is now pushing into continuous cybersecurity, including real-time monitoring, control validation, and application security testing.

That shift matters because a clean audit report starts aging with the next production deploy. A new integration changes access. A new AI agent touches customer data. A permission expands. The annual evidence package can still be accurate while the live system has already moved somewhere else.

Comp AI is not removing the people who carry responsibility. Lewis Carhart told TechCrunch that humans review and approve agent-drafted policies, and independent auditors still perform the actual examination. The product thesis is that software can do more of the continuous work without making approval or accountability disappear.

Grand Ventures also co-led Comp AI's $2.6M pre-seed with OSS Capital in 2025. The official disclosed rounds add to $36.6M. TechCrunch reports $37.5M in total funding, leaving an unexplained $900K gap. Precision is part of the compliance business too.

The new capital will fund product and engineering, expand continuous-security capabilities, and add people across Miami and New York. Comp AI has customer momentum and a public open-source core. It now has to prove the system can keep controls legible while software, vendors, permissions, and threats continue to change.

The founders left a million-user product because usage never became durable enough. This company is being built around an obligation enterprise buyers refuse to let their vendors ignore. The next stage belongs to the evidence Comp AI can keep current after the audit celebration ends.

Deeper Analysis 👇 https://devcuration.com/articles/comp-ai-raises-34m-continuous-security-compliance