The Machine Still Needs a Point of View

Curio

Curio just secured $2.5M in additional cost-shared funding from the U.S. Department of Energy's ARPA-E program, and the important word is not funding. It is additional.

ARPA-E first backed Curio's Closing the Cycle with NuCycle project with $5M in 2023. That work helped move the company's used nuclear fuel recycling process through laboratory demonstrations with DOE national labs. The new award adds another technical exam, not a victory lap: prove selected operations with actual used fuel and transuranics, then turn the results into engineering and regulatory work that can survive outside a slide deck.

Curio says it will pair the federal award with approximately $1M of company capital. At Pacific Northwest National Laboratory, researchers will test fluorination work designed to separate uranium and produce uranium hexafluoride. At Savannah River National Laboratory, the team will build and test electrowinning equipment intended to recover transuranics from molten salt as a solid metal mixture.

That is specialized chemistry with very ordinary business pressure. If a process cannot scale, clear regulatory scrutiny, and produce defensible economics, its cleverness stays trapped in the lab. This phase is meant to drag those questions into the open while there is still time to answer them.

CEO Ed McGinnis, Founder and Chairman Rabbi Yechezkel Moskowitz, Founder and COO Yehudah Moskowitz, and Chief Innovation Officer Vik Singh are pushing NuCycle toward a prototype laboratory and demonstration plant. Curio's stated target is capacity to process up to 4,000 metric tons of used nuclear fuel by the end of 2027. That is a target, not an operating result, and the gap between those two phrases is where the real work lives.

The market logic is hard to ignore. Commercial reactors leave most of their fuel's potential energy unused, while the United States stores a growing inventory of material that could contain reusable fuel and valuable isotopes. Curio is betting that better separation, safeguards by design, and a smaller process footprint can turn part of that liability into supply.

The skepticism deserves equal billing. Nuclear reprocessing has carried cost, security, and proliferation concerns for decades. ARPA-E support does not erase those risks. It funds the experiments, equipment, controls, and design work required to test whether Curio's version can handle them.

The Curio team now has another tranche of federal support behind the next phase. The sharper story is what comes next: less theater about nuclear waste, more measured proof about chemistry, controls, licensing, and economics.

AGent Energy

AGent Energy just raised an $11M Seed round to make an old piece of infrastructure do a new job: help keep the power grid steady before somebody has to build another plant and wait years for it to show up.

Spero Ventures and MassMutual Ventures co-led the financing. Intrepid Investment Management joined alongside returning investors CIV and Zero Infinity Partners, bringing AGent Energy’s reported total funding to $17M.

U.S. electricity demand is climbing, data centers are leaning hard on ERCOT and PJM, and grid operators need capacity faster than plants and permits can arrive. Backup generators already sit behind hospitals, factories, data centers, universities, water systems, and stores, waiting for emergencies.

AGent Energy turns that waiting room into a market. Sentinel hardware and the Command Platform connect, monitor, and dispatch backup generators when the grid is stressed. The host transfers to on-site power, grid demand falls, and the owner shares market revenue. It is capacity hiding in plain sight, with software deciding when it is useful.

The company says it has already brought more than 150 MW to market across ERCOT, MISO, and PJM after eight months of commercial sales, with a target of more than 1 GW over the next 12 months. That target is forward-looking, but the direction is clear: AGent Energy is trying to scale an existing fleet instead of waiting for an entirely new one.

AGent Energy was co-founded by CEO Stephanie Hendricks, President Françoise Parker, and Chief Development Officer Todd Krause. Hendricks brings experience from Voltus and NDimensional. Parker has spent more than 20 years in energy markets. Krause built teams across Voltus, Blue Pillar, and EnerNOC. CTO Bill Larkins earlier helped develop EnerNOC’s distributed-generation dispatch platform.

That team matters because this business is not merely software, hardware, or energy trading. It has to understand all three, then earn trust from asset owners, utilities, grid operators, regulators, and investors that rarely share one vocabulary, never mind one timetable.

The sharpest part of this round is what it says about infrastructure investing. The grid needs new steel, but it also needs better intelligence around the equipment already installed. Spero Ventures, MassMutual Ventures, Intrepid Investment Management, CIV, and Zero Infinity Partners are backing the layer that can identify, monitor, and monetize that overlooked capacity.

AGent Energy now has $11M in new capital to push that model further. The grid’s next meaningful resource may not arrive on a construction schedule. Some of it is already behind the meter, waiting for the right agent.

Aligned Marketplace

Aligned Marketplace raised an additional $20M Series A led by Venrock, bringing its disclosed funding to $31M. The money matters, but the more interesting part is what the company put next to the financing: a measured employer outcome.

In a program launched in January 2025, engaged members at a Fortune 500 company recorded 12% lower total cost of care than a risk-matched national benchmark, equal to roughly $96 per member per month. Aligned says Accorded independently measured the comparison.

That is sharper than another deck full of addressable-market math. Healthcare has never lacked ambitious projections. It has lacked models connecting access, clinical experience, and employer economics without asking benefits teams to rebuild the plumbing.

Aligned is designed around that constraint. The company gives self-insured employers and third-party administrators one path into a curated marketplace of independent advanced and direct primary care providers, spanning more than 3,000 in-person clinics across all 50 states.

Members get care at little or no cost. Independent doctors can serve distributed workforces without surrendering their practices to a hospital system, while employers get fees tied to engagement and outcomes instead of another bill that arrives whether anything improves or not.

The year-one data gives that alignment some weight. Aligned reported that 70% of engaged members were high risk, members averaged 4.7 visits with their Aligned doctor and 1.6 fewer visits elsewhere, and 94% said they could get care when needed.

Preventive care moved too. Aligned reported that mammography rates roughly doubled and colonoscopy rates nearly tripled after the first visit. Those numbers belong to one employer program, not every employer in America, but that is the right kind of proof to put under pressure.

Venrock leading the Series A adds a healthcare investor with an operating view of the category. Partner Bob Kocher, who sits on Aligned Marketplace's board, framed the bet around a national network that can make primary care more accessible while lowering costs for employers and patients.

Founder and CEO Patrick Nelli and the Aligned founding team are using the round to reach more employers and TPAs while expanding into value-based specialty care. Primary care can identify risk early, but the economics become more interesting when the downstream referral path also rewards value.

Aligned Marketplace now has additional capital to expand the model, but measurable outcomes are what turn a healthcare thesis into an operating system employers can trust.

Riven Systems

Riven Systems just emerged from stealth with a $7.9M pre-seed round to attack a problem the modern economy keeps pretending is a spreadsheet problem: critical minerals processing is chemistry, and chemistry does not care how persuasive the policy memo sounds.

The New York company is building an integrated AI platform for industrial chemistry. Its approach combines autonomous laboratories, computational modeling, and AI to design, de-risk, and operate mineral separation processes. The first target is rare earths, where the United States has resources but remains short on domestic processing capacity.

Mining gets the headlines. Processing decides whether ore becomes a usable input for defense systems, energy infrastructure, semiconductors, and advanced manufacturing. If the chemistry cannot separate it at the required purity, the supply chain is still somebody else’s business.

CEO Kurt (Chip) Breitenkamp, Ph.D., brings experience turning advanced materials science into products at scale, including prior leadership at NanoGraf. CTO Orion Archer Cohen brings computational chemistry and AI experience, including work on TorchSim and the Materials Project. One founder understands how laboratory discoveries survive contact with manufacturing. The other builds the computational machinery to find better chemistry faster.

Cerberus Ventures led the round. Scout Ventures, Alumni Ventures, Lightscape Partners, and CLAI Ventures participated, while New System Ventures served as Riven’s incubating investor. The capital will support the platform, technical hiring, and work on novel extractants and feedstock-specific separation systems.

Riven also received a U.S. Department of Energy Genesis Mission award and is collaborating with Berkeley Lab, Los Alamos National Laboratory, and UC Berkeley. That public-private mix carries weight because critical-minerals infrastructure has long timelines, national-security implications, and technical risk that venture capital alone does not dissolve.

The sharper lesson is that industrial AI will not look like another chat window wearing a hard hat. It will live inside laboratories, models, sensors, and plants, tightening the loop between experiment and deployment. Riven is betting that America does not need to reproduce four decades of foreign processing development one facility at a time. It needs a faster way to learn the chemistry.

A $7.9M pre-seed round does not rebuild a supply chain. For Kurt (Chip) Breitenkamp, Orion Archer Cohen, and the Riven Systems team, it can fund the machine that learns how.

Quarki Technologies

Quarki Technologies has secured pre-seed backing from India Accelerator, putting capital behind industrial optimization when the number of possible decisions gets stupidly large.

The amount was not disclosed. A precise number without confirmed disclosure is not intelligence. It is decoration wearing a spreadsheet costume. What is established is the investor relationship, the pre-seed stage, and Quarki's place in Finvolve's H1 2026 portfolio.

Quarki is building quantum-inspired solvers for combinatorial optimization. In normal language, the company is working on systems that help organizations choose better answers when logistics, supply chains, resource allocation, and finance produce more constraints than conventional planning can comfortably digest.

Quantum headlines often arrive dressed like science fiction, then leave operators wondering what they can actually buy. Quarki's pitch is narrower and more useful: borrow ideas from quantum mechanics, combine them with modern computation, and attack expensive decision problems without waiting for fault-tolerant quantum hardware to become office equipment.

The company was incorporated in Chennai in 2025 and is incubated at IIT Madras Research Park. CEO Sanyam Parashar, Chief Scientist Dr. Rajesh Narayanan, CTO Dr. Shashikant Kunwar, and founding member Dr. Vishnnu PK, who leads quantum algorithms, make up the core team.

India Accelerator identifies Quarki as a portfolio company, while Deepak Sharma identifies himself as a pre-seed investor and board observer. The bet is not on quantum as a fashionable noun. It is on a team trying to turn hard mathematics into operating leverage.

The market test is practical. Can Quarki produce answers fast enough, fit real workflows, and prove the savings beat adoption costs? Optimization buyers pay because routes get shorter, assets work harder, schedules stop fighting reality, and decisions improve under pressure. Customers, revenue, valuation, deployments, and use of proceeds remain undisclosed.

The next useful layer of quantum technology may not arrive as a glowing machine that changes everything overnight. It may arrive as software that quietly removes waste from decisions companies already make every day.

For Sanyam Parashar, Dr. Rajesh Narayanan, Dr. Shashikant Kunwar, Dr. Vishnnu PK, the Quarki Technologies team, India Accelerator, Finvolve, and Deepak Sharma, the funding is the headline. Turning constraint-heavy complexity into measurable business value is the assignment.

Vetspire

Vetspire just received a significant growth investment from Battery Ventures, and the story is bigger than another healthcare software check.

Veterinary practices run on clinical judgment, but the machinery around it is relentless. Records, scheduling, billing, inventory, communications, and reporting all compete for people who would rather be caring for patients. Vetspire puts those functions into one cloud-based system built around medicine first.

Plenty of vertical software starts with administration and eventually discovers the professionals forced to use it. Vetspire started with the clinician's day. Its platform combines practice management with AI-assisted documentation, patient summaries, specialty workflows, and operational tools for independent clinics and large multi-site groups.

The company says its AI Suite can save clinicians up to 90 minutes per day. It also reports that 58% of respondents in research with INSIGHT2PROFIT considered AI Scribe a must-have. Those are company-reported figures, but they point at the right pressure: less time fighting software means more capacity for care.

Battery Ventures is stepping in with a thesis that fits the category. General Partner Chelsea Stoner has spent years backing vertical software and healthcare IT, and Battery says Vetspire can grow as an independent provider for practices ranging from large networks to smaller clinics. The investment amount and terms were not disclosed.

Leadership is changing with the capital. Zachary Seely, who previously led acute-care facilities software company FSI, joins as CEO. Joseph Mazzarella continues as President. Co-founder Sam Ginn led Vetspire from 2018 through 2024 after building it from a firsthand view of after-hours veterinary documentation.

Pathway Vet Alliance acquired Vetspire in 2021 before becoming Thrive Pet Healthcare. The new investment does not spell out the ownership structure, so the cap table remains unclear. The direction is easier to read: position Vetspire as infrastructure for the whole market without forcing every customer into somebody else's operating model.

Veterinary software is becoming a test of whether AI can disappear into real work. The winners will be the systems teams trust during an emergency, at checkout, and at the end of a long shift when one more screen feels like one too many.

For Zachary Seely, Joseph Mazzarella, Sam Ginn, Chelsea Stoner, and the Vetspire and Battery Ventures teams, the capital creates a larger runway. The harder assignment is making sophisticated software feel like less work.

Bluehill.VC

Bluehill.VC just closed its maiden frontier-tech fund at more than ₹400 crore, roughly $42.1M using the stated conversion. The vehicle was built around a ₹350 crore base target plus a ₹50 crore greenshoe option. In a fundraising market where plenty of decks get polished and far fewer funds get finished, final close is doing real work here.

Co-founders and General Partners Manu Iyer and Sridhar Parthasarathy built the LP base across institutions, UHNIs, family offices, and founders. That mix matters. Frontier technology asks investors to underwrite engineering risk, long validation cycles, regulation, manufacturing, and commercialization before the revenue chart arrives.

Bluehill.VC is not shopping for thin software wrappers wearing a lab coat. Its thesis is IP-led technology across space, defence, semiconductors, energy, electric vehicles, advanced materials, robotics, biotechnology, manufacturing, and industrial systems. These are markets where a prototype can be brilliant and still have expensive years between proof and scale.

The fund is designed to back companies from seed through Series A, with initial checks generally ranging from $750K to $2M and larger follow-on capacity for selected companies. Portfolio examples include EtherealX, Zebu Intelligent Systems, Helex, optoML, and Sophrosyne Technologies. Different problems, same bet: serious technical advantage can become serious company value when capital is patient enough to let the work mature.

The timing makes the close more interesting. Moneycontrol reported in July that only 14 of 99 India-focused VC funds launched in 2023 and 2024 had completed fundraising. LP money still exists, but it has become pickier about who gets it, how focused the thesis is, and whether the manager can survive the distance between an exciting first close and a finished fund.

The business lesson sits inside the announcement. Specialization is no longer a clever paragraph in a pitch deck. It is an operating system. Bluehill.VC had to translate a frontier-tech thesis into an LP coalition, then turn that coalition into more than ₹400 crore of deployable conviction. Now the harder scorecard begins: selection, reserves, technical judgment, and patience through the parts of company building that refuse to fit a quarterly narrative.

Manu Iyer, Sridhar Parthasarathy, and the Bluehill.VC team have finished the raise. The capital now has a job: find Indian founders building defensible technology from first principles and give them enough runway to prove that hard science can become durable enterprise value.

Micron Ventures

Micron Ventures just put $250M behind a simple idea with expensive consequences: the next generation of AI will be shaped as much by memory and storage as by models.

The new Micron Ventures Paradigm Fund is the firm's third and largest fund. It will invest across the full AI technology stack, from model architecture and data infrastructure to compute, enterprise applications, robotics, and other forms of physical AI.

AI is moving from systems that generate answers toward systems that reason, act, and interact with the physical world. Every step adds pressure below the glossy demo: more data movement, more bandwidth, more power, and tighter latency.

Micron Technology knows where those demands land. The company builds the DRAM, NAND, and NOR memory and storage products that carry AI workloads from cloud infrastructure to edge devices. Micron Ventures gives the parent company a seat closer to the architectural arguments before they harden into product requirements.

The Paradigm Fund will focus on four connected domains. Model architecture covers the systems and data infrastructure shaping future compute needs. Compute includes in-memory processing, next-generation networking, and data-center efficiency. Enterprise applications bring AI into complex industries such as semiconductor design and manufacturing. Physical AI moves the fight into robotics and new device form factors.

This is corporate venture capital working as a listening post. Investment returns matter, but the strategic intelligence can be just as valuable. Back the people changing how AI systems are built, and you learn earlier what those systems will demand from memory, storage, networking, and power.

Micron Ventures has been building toward this position for years. Fund I was announced in 2018 with up to $100M focused on AI. Fund II arrived in 2022 with $200M for broader deep-tech investments and is still deploying capital. Add the new $250M vehicle, and Micron Ventures' total capital commitment reaches $550M.

René Hartner, Micron's Vice President of Corporate Development, framed the opportunity around AI paradigm shifts converging on high-performance memory and storage. Micron is not waiting for the next infrastructure pattern to show up in a procurement forecast. It is paying to see the pattern while startups are still inventing it.

The third fund puts another $250M behind that strategy. The sharper question now is which founders can turn its broad mandate into products that change what AI can do and what the hardware beneath it must become.

Vertical Insure

Vertical Insure just gave vertical SaaS companies a sharper answer to a question every operator eventually hears: once subscriptions and payments are working, where does the next meaningful revenue stream come from?

Updata Partners VIII led a growth investment in Vertical Insure, with existing investors Rally Ventures and Dundee Venture Capital participating. The amount and valuation were not disclosed. This is the first investment from Updata Partners VIII, and General Partner Jon Seeber will represent Updata on the board.

Vertical Insure lets platforms place tailored insurance inside registration, checkout, billing, and booking. Its multi-carrier API designs, prices, and administers coverage at the point of need. That keeps buyers out of disconnected broker processes and lets software platforms add revenue without pretending they became insurance carriers overnight.

The model is moving beyond theory. Vertical Insure says it supports more than 100 platforms across millions of transactions, spanning sports, travel, events, education, construction, rentals, associations, and other specialized markets. Named partners include Blackbaud, LeagueApps, Togetherwork, SportsEngine, and Daxko. Updata says the business is growing more than 100% year over year while staying capital efficient.

Founder and CEO Brock Noland saw the opening after a low-value pontoon policy proved too expensive for a broker to sell efficiently and too awkward for a buyer to purchase. CTO Greg Blasko helped turn that insight into software infrastructure built for repeatable, embedded distribution. Vertical Insure previously announced a $4M seed round co-led by Rally Ventures and Dundee Venture Capital in 2022, followed by another $2M led by Greenlight Re Innovations in 2023.

The new capital will accelerate go-to-market execution, expansion into additional markets and verticals, and growth of the product and carrier network. Vertical software is running into a basic economic truth: customers do not need another tab, another login, or another upsell wearing a fake mustache. They need relevant products placed inside the workflow at the exact moment those products become useful.

Brock Noland, Greg Blasko, Jon Seeber, and the Vertical Insure, Updata Partners, Rally Ventures, and Dundee Venture Capital teams are now pushing that model into its next stage. The bigger signal is not that software companies can sell insurance. It is that the best vertical platforms are becoming distribution systems for specialized financial products, and the infrastructure behind that shift deserves the same attention as the applications themselves.