Before They Say No

Proxy Foods AI Raises $6M Seed for Food R&D Platform

Proxy Foods AI just announced a $6M seed round, and the real story is not another startup attaching AI to an old workflow. It is a bet that food R&D can finally move from trial, error, and institutional memory trapped in spreadsheets toward predictive science.

Ted Leonsis led the round, with lead participation from Robert G. Hisaoka and SWaN & Legend Venture Partners, represented by Anthony Nader and Fredrick Schaufeld. The funding will deepen food-science, regulatory, and commercialization capabilities while supporting global expansion.

Developing a food product is a hostile little math problem disguised as lunch. Taste, texture, nutrition, cost, shelf life, sourcing, labeling, process constraints, and regulation all have opinions. Change one ingredient and the formula can behave like a committee meeting with no agenda.

Founder and CEO Panos Kostopoulos and CTO Petros Mandalis are building Proxy Foods AI to give R&D teams an agentic platform for formulation, reformulation, optimization, sensory work, regulatory analysis, and commercialization. The company says its system is built on more than 1M formulations and 5M product records.

Proxy Foods AI reports customers and ecosystem users including Barilla, Gefsinus, MISTA, and CAVA. It also says customers have reduced product-development iterations by as much as 80%, cut cost of goods by 2% to 10% in reformulation work, and compressed prototype timelines from months to days. These company-reported outcomes point to fewer expensive bench cycles before production money gets burned.

The company also says it worked with Stanford University and Food System Innovations on formulation research, while an Aperio Global partnership aims to add Zero Trust and quantum-resilient security.

Proxy Foods AI announced $2.3M in seed funding in 2024. The new release describes a $6M seed round but does not say whether that figure is entirely incremental or includes the earlier close. Precision matters, especially when the product itself is selling precision.

For investors, the signal is vertical AI moving into workflows where domain data, regulation, and physical-world testing create a real moat. For operators, the lesson is simpler: AI earns its seat when it reduces iterations, protects margin, and helps a team make a better product without pretending expertise can be automated away.

Panos Kostopoulos, Petros Mandalis, Ted Leonsis, and the Proxy Foods AI team now have more capital behind that thesis. The next test is how quickly predictive formulation can become normal infrastructure for the companies feeding everyone else.

Shelfmark Raises $3.5M Seed for Manufacturing Physical AI

Shelfmark just raised a $3.5M seed round to make some of manufacturing's fastest production lines smarter. The real story is what happens when a line can see a defect, connect it to the conditions that caused it, and help operators prevent the next one before good material becomes expensive scrap.

Armory Square Ventures led the round, with Grand Ventures, Hyde Park Angels, Argon Ventures, and Cultivation Capital participating. The financing brings Shelfmark's total capital raised to approximately $5M, according to the company.

Founded in 2022 by Pat O'Donnell, CEO, Shelfmark builds a managed Physical AI platform for continuous-flow manufacturing. Its initial markets include industrial films, decorated apparel, webbing, and building components, where materials move fast and production variation can make manual inspection unreliable.

The platform combines in-line cameras, spatial and environmental sensing, edge computing, deep-learning vision models, and software that turns quality data into production intelligence. Shelfmark manages the hardware, models, tuning, and ongoing performance. Manufacturers do not need another science project beside the line. They need a system that works while orders are moving.

Shelfmark says the technology was developed through work with 40 manufacturing facilities and has produced a 90% pilot conversion rate. The company also reports up to 90% waste reduction, 99.5% defect-detection accuracy, inspection labor costs cut in half, and returns of up to 7x in customer deployments. Those are company-reported results, not universal promises, but they put useful numbers behind a category that can disappear into abstract industrial AI claims.

One example makes the thesis tangible. Shelfmark says its system connected defect rates to changes in ambient humidity at a customer site. After humidity controls were implemented, the manufacturer's defect rate fell 50%. Catching bad output is valuable; understanding why it happened starts changing the economics of the entire run.

The new capital will support sales and marketing, more deployments, and continued development from detection into root-cause analysis, prediction, prevention, and guided corrective action. That progression is the difference between inspection as a cost center and production intelligence as operating leverage.

Pat O'Donnell and the Shelfmark team now have more capital behind that progression, with Armory Square Ventures, Grand Ventures, Hyde Park Angels, Argon Ventures, and Cultivation Capital backing the next phase. The bigger question for manufacturers is no longer whether AI can see the factory floor, but whether the systems being deployed can learn enough from every run to make the next one better.

Outside Interactive Secures $50M Credit Facility

Outside Interactive just secured $50M from Vector Velocity, and the key word in that sentence is not “million.” It is “secured.”

This is a senior secured credit facility, not an equity round dressed up for the applause track. Vector Velocity, the direct-credit strategy of Vector Capital Management, is lending into a company that has spent years assembling one of the broadest ecosystems in outdoor media and technology.

Outside is no longer just the magazine people leave on a coffee table while planning a trip they swear they will take next summer. The platform connects media, mapping, activity tracking, travel software, event services, subscriptions, and live experiences across 25 brands.

The company reports more than 70M monthly consumers, over 100M registered users, more than 1M paying subscribers, and more than 300M outdoor consumers reached annually. Those numbers explain why the financing makes sense. They also explain the real assignment.

A wide audience is not automatically a connected business. A portfolio is not automatically a platform. The value appears when a reader becomes a mapper, a traveler becomes an event participant, a runner becomes a subscriber, and the same identity can move across those experiences without the company asking people to start over every time.

That is the bet Robin Thurston and the Outside team are making. The $50M gives Outside more flexibility to invest in the products, technology, and experiences that tie the ecosystem together. Pricing, maturity, valuation, and a granular spending plan were not disclosed, so anyone pretending to know the capital structure from a headline is doing fan fiction with a calculator.

Vector is underwriting a different kind of media story. Vector Velocity targets senior secured loans to middle-market technology and tech-enabled businesses with durable positions and recurring revenue. Outside sits at an unusual intersection: trusted editorial brands on one side, utility products and subscriber behavior on the other, with events and travel creating more ways for the relationship to become real.

The takeaway is not that outdoor media found money. It is that an outdoor participation platform found credit willing to underwrite the connective tissue between content, software, commerce, and real-world activity.

Now Outside has to prove the ecosystem is more valuable together than it is as a collection of beloved brands. That means turning reach into repeat use, registered users into paying relationships, and a collection of outdoor interests into one product experience.

Sapiom Raises $35M Series A to Build the AI Agent Runtime Layer

Sapiom just raised a $35M Series A to solve the part of the AI agent boom nobody can finesse with a cleaner demo: production.

Dragonfly led the round, joined by Accel, Gradient, Coinbase Ventures, Operator Collective, Formus Capital, and VanEck Ventures. Okta Ventures, Menlo Ventures, Anthropic, and Array Ventures continued their support. Sapiom now reports $50M in total funding, and Haseeb Qureshi, Managing Partner at Dragonfly, is joining the board.

That investor mix fits because Sapiom sits where AI infrastructure, developer tooling, and programmable money argue over who owns the bill. Founder and CEO Ilan Zerbib co-founded Earny, then led payments engineering work at Shopify as Shop Pay scaled. The lesson traveled well: autonomous software is only as useful as the systems that let it act, pay, recover, and explain itself.

Sapiom began with financial access for agents. The company has now widened the frame into three products. Router matches each model call to an allowed path based on cost, quality, latency, reliability, and policy. Agent Studio helps builders create and inspect agents as real code. Runtime supplies the sandboxes, memory, secrets, tools, retries, and traces required after the demo glow wears off.

Gartner has forecast that more than 40% of agentic AI projects will be canceled by the end of 2027, citing cost, unclear business value, and weak risk controls. The market does not need more agents that look brilliant for eight minutes. It needs agents whose economics can survive a finance meeting and whose actions can survive an audit.

Sapiom says it processed more than 270M transactions in its first six months and now supports more than 100,000 agent runs per day. The company also reports that one customer cut inference costs by 75%, while another moved from roughly $1.2M to roughly $100K in monthly inference spend. Those are company-reported numbers, but they describe the exact pressure this round is underwriting.

There is a useful business lesson here for every founder building above the model layer. Intelligence is becoming abundant. Controlled execution is not. Once an agent can choose a model, call a tool, spend money, retry a failed step, and keep moving without a human standing over it, the operating layer stops being plumbing and starts becoming the product risk.

Sapiom and Ilan Zerbib have turned an unglamorous constraint into a serious infrastructure thesis. The round is also a wager that agent companies will be measured by what runs reliably, what each task costs, and what operators can prove after the run. How many AI roadmaps still have no honest answer to those three questions?

Avatar Robotics Raises $6.5M Seed for Industrial Humanoids

Avatar Robotics raised a $6.5M seed round led by AlleyCorp, with Defy VC and Headline joining the table. That is a compact round in a robotics market addicted to giant numbers, but Avatar is making a useful argument: the robot does not need a stadium entrance if it can show up for a shift.

Founder and CEO Colin Webb is building for the industrial work that fixed automation keeps leaving behind. Sorting, packing, kitting, picking, and carting are variable jobs where the SKU changes, the packaging changes, and the machine still has to earn its spot before lunch.

Avatar combines humanoid robots, remote human operators, and an AI autonomy pipeline. When the software meets an edge case, an operator can step in. Every shift generates real-world data, allowing the machine to become more capable without forcing the customer to redesign the facility around it.

Robotics has spent years looking spectacular in controlled demos and far less time surviving a live warehouse. Productiv moved two Avatar humanoids into production at its Dallas operation in November 2025. In March, Productiv reported 600 to 1,000 units per hour over an eight-hour shift, or 4,800 to 8,000 picks per day, with about 150,000 repetitions per robot each month.

Those are customer-reported numbers. They are still more valuable than another video of a machine performing one beautiful task where nothing unexpected happens. Real operations are where packaging bends, priorities change, and somebody asks the robot to handle a new product 10 minutes before the line starts.

Avatar sells the work as a service. The company supplies the hardware, and customers pay an hourly rate tied to output. That lowers the capital barrier and shifts the conversation from owning a robot to buying dependable capacity. For operators, it turns robotics from a science project into an operating decision.

Colin Webb co-founded Sauce Technologies, built AI tools for restaurant operators, and sold the company to Linked Eats in 2024. Now the same instinct is pointed at physical work: find the repetitive process, build intelligence around the exceptions, and make the economics legible to the buyer.

AlleyCorp leading, with Defy VC and Headline participating, signals a bet on human-in-the-loop autonomy as a bridge to broader industrial adoption. Full autonomy can remain the long game. Customers have shifts to staff tonight.

Colin Webb and the Avatar Robotics team now have $6.5M behind that thesis. The more important number will be how many messy production hours these machines can complete when nobody is filming.

Delightree Raises $25M Series A for AI Franchise Operations

Delightree just raised $25M, and the money is chasing a problem hiding in plain sight: companies with the most locations often have the least visibility into what is actually happening inside them.

Innovius Capital led the Series A, with Accel, Timber Grove Ventures, and Emergent Ventures participating. Announced August 4, the round gives Delightree fresh capital to expand engineering and product while accelerating its AI operating system for franchise and multi-unit brands.

Most software in this market records work after it happens. Delightree is trying to help run the work itself by connecting training, audits, compliance, communications, daily tasks, and new-location launches in one operating layer.

For frontline employees, that means answers from approved company knowledge instead of a binder. Franchisees can see where execution is slipping. Franchisors can find a delayed launch or unresolved audit issue before a monthly report turns it into ancient history.

The company says revenue has grown nearly 20x in just over two years, and the platform now spans more than 6,000 locations nationwide. Delightree also reports that customers can open locations two to three months faster while frontline teams reclaim up to 30% of their time.

This is not AI sprinkled over a legacy dashboard so the sales deck can wear a new suit. Delightree is betting on agentic operations: software that surfaces problems, identifies the next action, assigns ownership, and tracks the fix while humans stay in control.

The opportunity reaches across restaurants, gyms, clinics, salons, retailers, and home-service networks that live or die by consistency, yet often run on spreadsheets, inboxes, disconnected tools, and institutional memory. Scale exposes every seam.

Tushar Mishra, CEO and co-founder, is building around that execution gap. Madhulika Mukherjee, co-founder, is pushing the product toward a model where AI helps turn approved standards into action across the network, not another place to store documents and forget them.

Innovius is backing the move from systems of record to systems of execution. Accel and Emergent have supported Delightree since its $3M seed in 2020, while Timber Grove says it first invested in 2024. That mix of new conviction and returning capital is a useful signal.

Tushar Mishra, Madhulika Mukherjee, and the Delightree team now have more capital behind the execution thesis. The question is no longer whether multi-unit brands need better software. It is how quickly AI can move from answering questions about the playbook to helping every location run it.

Hadrian Raises $1.37B Series D for Defense Factories

Hadrian just raised a $1.37B Series D at a post-money valuation just under $8B. That is Wall Street looking at the defense industrial base, hearing metal grind against a production deadline, and deciding the factory floor deserves growth-stage capital.

JPMorganChase's Strategic Investment Group anchored the round through its Security and Resiliency Initiative. WCM Investment Management, Washington Harbour Partners, Valor Equity Partners, 137 Ventures, and Baillie Gifford joined as co-leads. This is a coalition underwriting capacity, not one fund chasing a shiny defense-tech story.

Founder and CEO Chris Power started Hadrian in 2020 around a hard truth: America can design extraordinary systems, then spend years waiting on the parts. Hadrian attacks that gap with automated factories, trained operators, and Opus, its factory-autonomy software. The company provides precision components, embedded production cells, supply-chain services, and dedicated factories.

Hadrian now has nearly 3M square feet across four sites, with a new Los Angeles headquarters and a San Francisco engineering and R&D hub planned. Its customers include Lockheed Martin, RTX, and Anduril. Lockheed Martin also confirmed a partnership to deploy Hadrian's machining and inspection technology inside missile production.

Then there is Factory 4 in Cherokee, Alabama. The U.S. Navy says the 2.2M-square-foot site will mass-produce components for Virginia-class and Columbia-class submarines. The project combines $900M in Navy investment with $1.5B in private capital and could create up to 1,000 jobs. That financing is separate from this Series D, but capital is clearly meeting a production problem measured in years, ships, and qualified parts.

Hadrian raised a $260M Series C in July 2025. In January 2026, new financing valued the company at $1.6B. Seven months later, the post-money valuation is nearly $8B. The headline is the markup. The harder story is the burden attached to it.

Hadrian is not being valued like a clever machine shop. It is being valued like infrastructure. The next proof point is whether Opus, the workforce model, and the factory system can repeat across programs without quality, qualification, or delivery discipline slipping.

Chris Power and the Hadrian team now have an enormous amount of capital behind an equally large assignment: turn American reindustrialization from a conference-panel consensus into dependable output. If Hadrian can make capacity repeatable, its most important product may be a manufacturing model the country can finally scale.

Verix Makes Syntasa Its First Platform Investment

Syntasa just became the first platform investment for Verix Equity Partners, and the deal says more about enterprise AI than the undisclosed price tag ever could.

The strategic investment keeps founder and CEO Jay Marwaha in charge with significant ownership alongside Verix. That continuity matters. Syntasa has spent more than a decade building a sovereign AI and data platform for organizations that cannot treat data custody like a casual checkbox. Government agencies, classified environments, and regulated enterprises do not get to shrug when a model call crosses the wrong boundary.

Syntasa's answer is architectural: bring the models and agents to the data, keep execution inside infrastructure the customer controls, and make every pipeline, decision, and action auditable. The customer keeps the data, the models, and the authority.

That premise becomes more valuable as the market moves from chatbots that suggest to agents that act. Model access is becoming abundant. Trusted context, governed execution, and a provable chain of responsibility are not. When an agent can take action, the security perimeter stops being an IT detail and becomes part of the product.

Jay Marwaha remains the steady hand on the business side, while CTO Shawn Zargham has been building the architecture underneath that promise. Verix co-founders and Managing Partners Jeff Hart and Sean Battle are backing a company that sits directly inside the firm's national-security thesis: founder-led technology, mission-critical operations, and infrastructure that must work when external dependencies are unacceptable.

Verix launched in June and chose Syntasa as its inaugural platform two months later. That is a sharp opening statement. The firm is starting with the less glamorous layer that determines whether AI can be trusted with sensitive work at all.

The growth plan follows the thesis. Syntasa and Verix intend to accelerate R&D in sovereign AI, agentic workflows, and secure deployment, expand go-to-market capacity across national security and commercial enterprise, and pursue targeted acquisitions across the sovereign data and AI ecosystem. Terms were not disclosed, so there is no honest check-size flex to perform here.

The sharper signal is control. AI adoption will not be decided only by who has the smartest model. It will also be decided by who can connect intelligence to consequential data without surrendering custody, governance, or accountability.

Jay Marwaha, Shawn Zargham, and the Syntasa team now have Verix behind the next stage of that thesis. Verix did not just make its first platform investment. It picked a thesis with teeth.

Solar Landscape Secures Up to $150M From CIP

Solar Landscape just secured up to $150M in development financing from Copenhagen Infrastructure Partners, and the headline is less about another clean-energy check than where the capital enters the machine.

The facility comes through CIP’s Green Credit Fund II and initially supports roughly 250 MWdc of mid- and late-stage distributed solar assets. It targets the expensive stretch before construction: equipment procurement, interconnection, offtake work, and sponsor development services. That stretch decides whether a project becomes infrastructure or remains a handsome spreadsheet with excellent intentions.

Solar Landscape has built its edge around commercial rooftops, turning underused real estate into distributed generation close to demand. Its vertically integrated model covers development, engineering, construction, ownership, and operations. Flexible early-stage capital gives that machine more room to move projects toward construction before timing friction starts collecting rent.

This is not a traditional venture round, and “up to $150M” is not the same thing as $150M wired on day one. It is a development financing facility built for forward-flow portfolio execution. Precision matters when the market treats every large number as interchangeable confetti.

The timing is sharp. Electricity demand is rising, the PJM grid is constrained, and long transmission timelines have made local generation more valuable. Rooftop solar cannot solve every grid problem, but generation that connects closer to load can move differently than remote projects waiting behind transmission queues.

Solar Landscape says the transaction follows a $600M senior debt facility, a $125M corporate revolving credit facility, and a $117M preferred equity investment. Together, the company reports raising more than $2.2B over the last three years. That stack tells the deeper story: sophisticated capital is financing different layers of the same deployment engine.

Solar Landscape gets capital aimed at throughput, not applause. The first portfolio represents about 250 MWdc, and the company says it has deployed more than 500 projects totaling over 750 MWdc.

Founder and CEO Shaun Keegan, COO Corey Gross, CFO Clayton Avent, CTO Paulo Perillo, and the teams turning rooftops into grid assets now have more capital behind that deployment engine. Reiner Boehning and Copenhagen Infrastructure Partners are backing a model that treats development velocity as a financial advantage.

Solar Landscape and Copenhagen Infrastructure Partners now have up to $150M committed to that thesis. The next proof point is how quickly those 250 MWdc move from development risk to operating power.

Lumilens Raises $700M+ to Scale AI Data Center Optics

Lumilens just pulled more than $700M into the connective tissue of AI infrastructure, pushing total funding above $900M and the company's valuation to $5.51B. The round is reported as Series C, but the more important label is an infrastructure reality check: buying GPUs is only half the job when those processors still need to behave like one machine.

Atreides Management, Bain Capital Ventures, Meritech Capital, Seligman Ventures, and Spark Capital co-led the financing. Mayfield, Qualcomm Ventures, Peak XV Partners, and others also participated.

Founded in early 2024, Lumilens builds photonic interconnects for scale-out and scale-up AI systems. Its lineup spans pluggable, near-package, and co-packaged optics, with silicon photonics, mixed-signal ICs, electrical-optical interposers, and optical systems pulled into one architecture. Lumilens is betting light can move data faster and more densely than copper as clusters grow.

Lumilens says its first scale-out product has completed qualification and is shipping into production AI data centers under a multibillion-dollar customer agreement. That customer remains unnamed, so restraint matters. Still, qualification and production matter more than a polished demo when a networking miss can turn a staggering GPU budget into very expensive furniture.

The company is led by founder and CEO Ankur Singla, alongside founders Samuel Liu, Ritesh Kapahi, and Dave Friedman, with Ted Schmidt serving as CTO. The group brings experience from Cisco, Juniper Networks, Meta, Lumentum, and Coherent Corp. This is not a team learning optical systems from a slide deck after investors discovered the category.

The financing will expand engineering, software, silicon photonics, process engineering, and high-volume manufacturing. Photonics wins do not matter if precision assembly stays slow, expensive, or difficult to repeat. Lumilens is pairing product design with robotics and automated manufacturing because scale is where elegant hardware stories meet a wrench.

Its May partnership with POET Technologies adds commercial texture: an initial $50M purchase order for EOI-based engines, with a conditional framework that could exceed $500M over five years if development, qualification, and manufacturing milestones hold. "Conditional" belongs in the sentence, but so does "$50M purchase order."

Ankur Singla, Samuel Liu, Ritesh Kapahi, Dave Friedman, Ted Schmidt, and the Lumilens team now have more than $700M behind the next stage of that build. This round says investors are no longer treating AI networking as background plumbing. The chips get the attention. The connective layer determines whether the infrastructure can deliver on the compute being poured into it.

HealthSnap's $25M Series B Backed Care-at-Home Infrastructure

HealthSnap closed a $25M Series B on February 21, 2024. The interesting part was not another digital health company discovering that patients own couches. It was a bet on what happens between office visits, where chronic conditions do damage and care teams have the least visibility.

Sands Capital led the round. Comcast Ventures, Acronym Venture Capital, and Florida Opportunity Fund joined as new investors. Asclepius Growth Capital, Florida Funders, MacDonald Ventures, and TGH Ventures also participated. HealthSnap said total funding reached $48.5M, and Scott Frederick of Sands Capital joined the board.

They backed a model built around Remote Patient Monitoring and Chronic Care Management, not a prettier portal with a motivational notification. HealthSnap combines connected devices, care navigation, EHR integrations, billing workflows, and population analytics so providers can manage chronic conditions beyond the clinic.

At the announcement, HealthSnap reported more than 150 provider organizations and health systems, more than 300% growth in patient programs during the prior year, support for more than 100 chronic conditions, and integrations with more than 80 EHRs.

Founded in Miami in 2015 by Samson Magid, Chase Preston, and Wesley Smith, HealthSnap still has its founding team in the operating picture. Samson Magid serves as CEO, Chase Preston as COO, Wesley Smith as Chief Science Officer, and Jared von Halle as CTO. Virtual care does not become durable by stapling software to a nurse call center. Clinical operations, integration, and reimbursement have to work as one system.

HealthSnap said the capital would expand its national clinical team, accelerate the product roadmap, increase access to care in the home, and begin building an AI-powered population health and analytics solution. The AI line gets attention. The stronger signal is the data foundation underneath it: longitudinal readings and clinical workflows tied to real provider operations.

Remote monitoring lives somewhere between an obvious idea and difficult execution. The technology can be simple. The service model, patient engagement, reimbursement, and clinical accountability are where the bill arrives.

HealthSnap's Series B suggested investors saw value in a company trying to own that stack. The round did not disclose a valuation, and it did not need one to make the point. Care is moving beyond the clinic. The winners will help health systems turn data into timely, repeatable action.

The HealthSnap team, its founders, leadership, and investors now have the next stage of that execution ahead of them.

Tower Arch Capital Recapitalizes Boxzooka eFulfillment

Tower Arch Capital just recapitalized Boxzooka eFulfillment, and the interesting part is not a number. The transaction terms were not disclosed. The signal is what Tower Arch chose to buy after evaluating more than 100 fulfillment opportunities: a 3PL that treats software, service, and brand protection as the same operating problem.

Boxzooka was founded in 2014 by Brendan Heegan for e-commerce brands that cannot afford generic fulfillment. Apparel, beauty, wellness, and premium consumer businesses live or die on details that disappear inside a standard warehouse pitch. Orders and packaging have to be right, while returns cannot turn customer loyalty into a hostage situation.

Boxzooka pairs proprietary Warehouse and Transportation Management Systems with high-touch omnichannel fulfillment, including reverse logistics, custom kitting, premium unboxing, and retail distribution. Its network spans four facilities across New Jersey, Pennsylvania, and Nevada, giving brands physical reach with real-time order visibility.

Tower Arch is bringing strategic guidance, governance, M&A support, and capital. Boxzooka says the partnership will accelerate technology investment, expand its logistics network, and improve the client experience. First Financial Bank provided transaction financing, while equity terms and valuation remain undisclosed.

The leadership transition matters, too. Brendan Heegan will remain an owner and become Chairman in an advisory role. Nick Pellegrino, president and COO at the announcement, becomes CEO. Bo Li remains CTO. That keeps founder context in the room while day-to-day leadership goes to an operator with 30 years of supply-chain experience.

Private equity has spent years searching for logistics businesses that are more than square footage with forklifts. Boxzooka's case is that fulfillment software and white-glove execution compound each other. Better visibility helps operators decide. Better execution protects the brand. That gives a growing merchant fewer reasons to replace its logistics stack mid-expansion.

Warehousing can look interchangeable from far away until a return is mishandled, a launch misses its window, or premium packaging arrives looking like it lost a bar fight. The valuable 3PL makes complexity boring for the customer without making the operation simplistic behind the scenes.

Brendan Heegan, Nick Pellegrino, Bo Li, and the Boxzooka team now have David Topham and Tower Arch Capital behind the next stage. The test is whether Boxzooka can scale its network, keep service personal, and make its proprietary logistics technology even harder to copy.

Triangle Raises $15M for Asia-Pacific Live Touring

Triangle Group just raised $15M, and the useful part is hiding in the split. This is not one oversized seed round. It is a $10M equity seed led by DSG Consumer Partners, with Golden Gate Ventures, the Oberoi Family Office, and angel investors participating, plus a $5M venture debt facility from Genesis Alternative Ventures.

That structure reflects what Triangle is building. Formerly known as Creation Collective, the company funds and operates tours across Asia and Australia, owns tour and festival P&Ls, routes artists across markets, and handles local promotion, production, and partnerships. Triangle is selling coordinated execution across borders.

Greg Hargrave and Zaran Vachha founded the Singapore company in 2024 around a stubborn market problem: global artists can have millions of listeners across Asia-Pacific and still face a fragmented touring map. Triangle's current footprint spans 18 cities in nine countries. DSG Consumer Partners says the predecessor business had already delivered more than 60 shows across 16 cities, demonstrating operating history rather than a map drawn after the money arrived.

The financing also arrives with an operator who has seen the machinery from inside the biggest room. Roger Field, formerly President of Live Nation Asia Pacific and CEO of Live Nation Australia & New Zealand, has joined Triangle full time as Executive Director and Operating Partner after serving as an advisor and investor. Capital can buy calendar. Experience can stop that calendar from becoming a regional collection of expensive surprises.

The larger signal is that the experience economy is becoming an infrastructure story. Audiences are global before touring systems are. Streaming can build demand market by market, yet monetizing it live still depends on local judgment, production discipline, and someone willing to own the P&L. Triangle wants Asia and Australia to operate like one touring region without treating their audiences like one interchangeable crowd.

Now comes execution. Triangle says the capital will support wider touring and festival infrastructure and future consumer services, including travel and hospitality packages. If Greg Hargrave, Zaran Vachha, and Roger Field can turn regional complexity into repeatable delivery, the company may change who gets to plan Asia-Pacific as a first-class tour leg.

The Triangle team now has $15M behind that expansion. The sharper story is a regional operator pairing equity, debt, and deep touring experience to make a fragmented market easier to enter without flattening what makes each market matter.

ReCode Therapeutics Gets New CFF Funding for Gene Editing

ReCode Therapeutics just put fresh capital and a new collaborator behind one of genetic medicine's least forgiving problems: getting the edit to the right cells in the lung.

On August 6, 2026, ReCode announced new funding from the Cystic Fibrosis Foundation to support a research collaboration with an unnamed gene-editing company. The amount was not disclosed, which matters because old funding numbers do not become new just because the headline changed.

The architecture is clean. ReCode brings its Selective Organ Targeting lipid nanoparticle platform, the partner brings gene-editing technology, and the teams plan to advance candidates designed to correct mutations in CFTR, the gene at the root of cystic fibrosis.

ReCode has spent years building that delivery layer. Co-founder Daniel J. Siegwart's work helped create the SORT platform, designed to direct genetic cargo beyond the liver and toward selected organs and cells. For cystic fibrosis, the lung is not just the destination. It is the obstacle course.

The company is also advancing RCT2100, an inhaled CFTR mRNA therapy intended to give lung cells working instructions for producing functional CFTR protein. Its Phase 2a study is fully enrolled, FDA Fast Track designation is in hand, and ReCode expects data in Q4 2026. That program is separate from the new editing collaboration, but together they show the strategy: pursue repeatable mRNA delivery now while building toward more durable correction.

The leadership change fits the moment. Heather L. Clark became CEO effective July 1 after leading ReCode's cystic fibrosis franchise and its Foundation relationship. Clark spent more than 20 years at Vertex Pharmaceuticals helping discover and develop five FDA-approved cystic fibrosis therapies. Shehnaaz Suliman moved from CEO to Executive Chair, maintaining strategic continuity while Clark takes the operating seat.

The Foundation's role goes beyond writing checks. It brings disease expertise, infrastructure, and a patient-centered filter.

The sharper signal is focus. ReCode is concentrating capital, delivery science, editing capability, and leadership experience on patients who remain underserved by current CFTR modulators. The new partner is unnamed, the commitment is undisclosed, and the editing program is at the research stage. Those are boundaries, not footnotes.

ReCode, Heather Clark, Shehnaaz Suliman, Daniel J. Siegwart, and the Cystic Fibrosis Foundation are keeping the hard problem in view. The next test is whether Q4 data and this collaboration can turn delivery precision into meaningful cystic fibrosis evidence.

Vedanta Biosciences Secures $60M for Phase 3 Trial

Vedanta Biosciences just secured $60M, but the structure matters as much as the headline.

The Cambridge biotech raised a $40M financing round led by existing investors AMR Action Fund and BNP Paribas Asset Management Alts, with roughly $18M arriving in Q1 and $22M in Q2. Another $20M came through Vedanta's existing contract with the Biomedical Advanced Research and Development Authority (BARDA).

That is private capital and public-health funding converging on one very specific job: get RESTORATiVE303 through its pivotal Phase 3 work and closer to a regulatory decision.

RESTORATiVE303 is testing VE303 for the prevention of recurrent Clostridioides difficile infection. VE303 is an oral live biotherapeutic made from a defined consortium of eight bacterial strains grown from pure clonal cell banks. In plain English, Vedanta is trying to turn microbiome medicine into a standardized pharmaceutical product, not a treatment dependent on variable donor material.

The business thesis sits inside that scientific distinction. A platform must become a reproducible product, survive a controlled trial, and produce data regulators can evaluate.

Vedanta says the Phase 3 study has crossed 80% enrollment across more than 150 sites in roughly 20 countries. Enrollment is expected to finish in the second half of 2026, with topline efficacy data targeted for the first half of 2027. Those dates remain company expectations, but they put the next proof point on the calendar.

The company also strengthened its board. Manos Perros, Ph.D., was appointed Executive Chairman, and Andrew Davis joined as an independent director. Both bring drug-development and strategic transaction experience.

Founder and CEO Bernat Olle is not selling a vague microbiome future here. The financing is tied to a late-stage program with a defined endpoint: recurrent C. difficile infection at Week 8 after treatment. That kind of clarity is rare enough in biotech to deserve attention.

The market lesson is simple. Capital gets more useful when the next question is sharper. Vedanta now has $60M of additional support, a pivotal trial beyond 80% enrollment, and a visible data window. The story from here will not be decided by funding theater. It will be decided by whether VE303 can convert a differentiated manufacturing thesis into Phase 3 evidence. Now the clinical data has to carry the argument forward.

Bernat Olle and the Vedanta Biosciences team now have AMR Action Fund, BNP Paribas Asset Management Alts, and BARDA behind this next stage. The hard part is still ahead, which is exactly why the financing matters.