What the Shortcut Still Owes

Punch

Learning who can approve a material order is part of learning a construction company. Punch wants that answer available before the new hire has to track down the person who knows.

Taylor Sandusky, Punch's founder and CEO, spent years managing construction projects. He is building around the knowledge that makes a job run: which supplier to call, which price applies, who signs off and what deserves the project manager's attention.

The Atlanta company, founded in 2025, announced $2M on October 15 from returning investor Gutter Capital, with customers and technology angels also backing the business. The announcement leaves the round type and valuation undisclosed. Dan Teran, Gutter Capital's managing partner, points to Sandusky's experience in the field as part of the investment case.

A superintendent can speak to Punch, send a photo or forward an email. The platform uses that update and the contractor's rules to prepare daily reports, create purchase orders, flag delays and assign follow-up. Estimating and accounting systems stay in place. Punch takes on the work connecting them.

That arrangement changes the introduction a contractor has to make to its software. The company has to explain how it buys, who gets involved and where a routine request stops being routine. A voice interface gets the information in. The contractor's operating knowledge gives it somewhere useful to go.

At Piedmont Drywall, general superintendent Dustin Gibbs reports cutting a daily report from 30 minutes to 3. Across reporting, material orders and payroll, Gibbs estimates saving about 2.5 hours a day. Those are his reported results, with a named person and a specific job behind them.

A separate company-published business case projects $178K in annual savings for a drywall contractor. The distinction matters: expected savings still have to survive a year of actual orders, invoices and decisions. A purchasing error avoided in the model needs to become an error avoided on the job.

Punch will use the money to grow engineering and customer teams, develop the platform and help more contractors adopt it. Customer support sits close to the product here. Someone has to help turn the way an experienced employee runs a project into instructions the next employee can use.

As Punch reaches more contractors, it will keep meeting companies whose best procedures have never been written down. Getting those procedures into the next material request gives the people already doing the work a little more room to teach it.

Acta Pharmaceuticals

A company founded in 2022 can inherit a scientific project older than its incorporation papers by two decades. Acta Pharmaceuticals is bringing that inheritance into the clinic, where the questions become personal.

The Boston biotech announced $24M in total seed financing on October 15. Co-founder and CEO George Alex built the company with scientific founder Dr. Rudolph Tanzi, whose research provides the foundation. CSO Kevin D. Rynearson carries the drug-design work forward from the laboratory of the late Dr. Steven L. Wagner.

Acta's lead candidate, GSM-779690T, comes from roughly 25 years of gamma-secretase modulator research. It is an investigational oral small molecule designed to change the balance of amyloid-beta peptides: reduce Aβ42, which contributes to plaque formation, while increasing shorter Aβ37 and Aβ38 forms.

Peptide production becomes the therapeutic target. Acta wants to intervene earlier in Alzheimer's disease and eventually investigate prevention, a goal that would carry the medicine into a very different relationship with people who may still feel well.

First comes the human evidence. The registered Phase 1 protocol studies safety, tolerability and pharmacology in healthy adults. Its July update still lists the study as not yet recruiting; Acta's October announcement describes advancing the program. Current dosing has not been independently established, and clinical benefit remains unproven.

Shareholders include Alzheimer's Disease Research Foundation, Dolby Family Ventures and Gamma Bio Investors, alongside family offices and mission-driven investors. Acta names no lead. The approximately $70M cited for earlier preclinical development belongs to the scientific history, separate from the $24M corporate seed total.

George Alex also negotiated the company's exclusive intellectual-property licenses from Massachusetts General Hospital. Turning academic discoveries into a drug company requires assembling the rights, scientific continuity and clinical operation that can carry them into people.

Acta has a second program developing combinations identified through human-cell models of Alzheimer's disease. Laboratory screening gives the team candidates to investigate; trials must establish what those candidates can do in patients.

For this team, the prevention ambition reaches back into the first safety study. The people asked to take a medicine before symptoms emerge would be placing years of trust in what Acta is beginning to learn now.

Mueon

Cooling gets a seat at the architecture meeting at Mueon Corporation. The Portland semiconductor startup is building Cubelets, modular 3D systems that bring compute, memory, interconnect, power delivery and thermal management into the same engineering decision.

That changes who has to agree before a denser machine becomes useful. Put more silicon into less space and the person moving heat has a direct say in how much computing the customer can actually use. The cold plate belongs in the product conversation from the beginning.

CEO and co-founder Wilfred Gomes and co-founder and CTO Jack Hwang bring years of Intel experience to that conversation. Gomes worked on 3D integration and Foveros packaging; Hwang brought semiconductor process expertise. They are applying that history to the physical organization of AI infrastructure, where a clever component still has to live with everything around it.

Mueon announced a $65M Series A on October 15, led by Maverick Silicon. Intel Capital, Geodesic Alliances Fund, Oregon Venture Fund, M-Ventures, Wellington Capital, IQT and Defined participated. The company reports more than $80M raised, following its $15.5M Seed in 2025. Maverick Silicon's Manish Muthal joins the board alongside Intel Capital's Srini Ananth.

The money supports engineering and operations hiring, broader system validation, production readiness and customer engagements. Mueon's current Thermal Systems Engineer role spans silicon, cold plates, trays and rack integration. That job description makes the funding tangible: somebody has to follow the heat through the whole machine.

Mueon is targeting up to 10X better power and cost efficiency in the same rack footprint. That remains a company target. Its announcement puts Cubelets on the path toward production readiness, with d-Matrix and Ampere describing technology collaboration. There is no disclosed shipment calendar or independent benchmark in the primary material.

The software team is part of the same undertaking. Mueon is building orchestration around familiar AI frameworks, so the proposed hardware has to meet developers where their work already runs.

For an AI infrastructure buyer, density comes with an operating relationship. The supplier has to make the compute, cooling and power choices work together, then make the resulting system usable. As Gomes and Hwang expand the team, the people responsible for those choices will be working across boundaries the architecture is trying to shrink.

Multiliquid

Saturday's card payments still need reserves. Multiliquid is building for the treasury team that would like those reserves earning a return until the moment somebody needs to spend them.

Putting a money market fund on a blockchain creates a new way to hold it. Getting cash out can still involve an issuer's redemption calendar, exactly when a payment business or trading desk has a calendar of its own.

Will Beeson, founder and CEO, knows the asset side of that relationship. Before founding Multiliquid in 2025, he helped develop institutional tokenization infrastructure and Wellington's ULTRA fund at Standard Chartered's Libeara. COO Angelo D'Alessandro built UniCredit's buddybank. CCO Michal Benedykcinski brings experience from Arca and Hedera.

Their Los Angeles business, developed by Uniform Labs, has closed a Seed round led by Polychain, with Strobe Ventures, CMT Digital, Metalayer Ventures, Generative Ventures and Circle Ventures participating. The October 15 announcement brings cumulative funding to $8M; the Seed amount was not disclosed separately.

Multiliquid connects tokenized assets and stablecoins in an atomic transaction on Ethereum and Solana. Both sides settle together, with issuer permissions carried through the transaction. The company says it supports funds from Wellington Management, VanEck, Janus Henderson and Fasanara, alongside stablecoins including USDC.

The useful commercial detail sits with the fund manager. An investor taking a secondary exit sells the position to a liquidity provider, so the token can stay outstanding and the fund keeps its assets under management. The departing holder gets access to cash while somebody else accepts the holding period and prices the exposure.

That gives the liquidity provider a business too: earn a spread and the underlying yield while supporting the exit. The software coordinates the exchange; the capital provider still has to want the position, at the price and under the rules available.

Multiliquid's Treasury documentation makes the practical limit visible. A stablecoin buffer supports immediate withdrawals, and redemptions beyond available stablecoins can return to the underlying issuers' timeframes. Treasury teams have a reason to read that paragraph as closely as the settlement-speed claim.

The funding supports more engineering and integrations. As more reserves move into tokenized products, the treasury team's next question gets very specific: how much USDC is available when the payments arrive?

Molagri

A pesticide gets a new opponent every time the pest changes. Molagri wants the people designing crop protection to stay in that conversation after the first formulation leaves the lab.

Co-founders Zaky Hassan, CEO, and Min Jin, CSO, studied human coronaviruses at the University of Toronto. They are bringing viral structural biology and protein biochemistry into an agricultural problem: how to kill a specific insect while leaving useful insects alone, then respond when resistance changes the target.

The Toronto startup announced $5.8M in combined pre-seed and seed funding on October 14. Convective Capital led the seed, joined by Y Combinator, Superorganism, Robinhood Ventures, Paul Graham and other angels. Julian Capital was the earliest backer and led the pre-seed. The total covers both rounds; the latest seed increment was not disclosed.

Molagri engineers non-infectious virus-like particles to recognize receptors in a pest's gut and deliver genetic instructions. The intended result is that the pest's own cells produce an insecticidal payload. Targeting and payload are modular, giving the scientists separate components to adjust as resistance evolves. The manufacturing plan uses yeast fermentation, putting repeatable production beside the biology.

There is a commercial relationship hiding inside that design. A grower buying protection for an orchard needs the supplier to keep understanding the insect. Molagri's ambition would carry molecular research into the life of a product, with the next useful formulation arriving before the old one becomes a bad bet.

Tree nuts and tree fruits are the starting markets. Molagri reports an initial grower letter of intent, while field efficacy, manufacturing costs and formulation stability remain to be established. A grower's interest gives the scientists a conversation to build on, with the evidence still being assembled at the bench.

That bench is getting more room. The company has expanded its Toronto lab, filed a provisional patent and is recruiting across insect biology, virology, protein targeting and VLP production. The funding buys people who can connect a promising molecular interaction to a reproducible result.

Regulatory review will help determine how quickly a redesigned product can reach a crop. Hassan says that approach is still being explored. The grower's calendar will eventually have to meet the scientists' ability to respond, with an insect continuing to evolve somewhere between the two.

Sharp Performance

The chief paying for Sharp Performance does not get a list of who called the coach. That missing list is part of what the department is buying.

Police officers, firefighters and paramedics can have support available and still hesitate to use it. A conversation carries different weight when the person listening also knows the people deciding your next assignment. Sharp puts the coach outside that chain of command, with a background the member can recognize.

CEO and co-founder Ben Curley and co-founder Andrew Sakmar, a former Green Beret, have built a marketplace around that relationship. Sharp's $18M Series A is led by Hidden Variable, with Andreessen Horowitz, Commonweal Ventures, Wolfson Group, FJ Labs and the Ford Foundation participating. The company reports $30M in total funding.

There is software here, but the supply is people. Former officers, firefighters, paramedics and special operators bring experience that cannot be added with another onboarding screen. Sharp says each coach is recruited, vetted and trained individually. A larger department network gives the company more room to build that roster and find a useful fit.

Members work through 30-minute phone or video sessions, choose their cadence and can change coaches. The agenda can follow a promotion, a return from leave, a difficult shift or the routines waiting at home. The app keeps coaching, practical exercises and department resources accessible between those conversations.

Sharp says more than 110 departments across 18 states have adopted the service, with members booking over 16,000 sessions. Those are company-reported usage figures. They show members booking time for a conversation, while leaving the harder questions about lasting outcomes open.

Departments receive aggregate reporting, with safety and legal limits to confidentiality. Coaching stays separate from diagnosis and treatment, alongside the clinical and peer-support resources a department already has. Keeping those boundaries clear is work the product has to keep doing as it grows.

The department pays; members are not billed, and there is no session cap. Spouses and retirees can have their own coach, extending the relationship beyond the person on the duty roster.

The financing supports more departments, improvements to the app and additional products. Each rollout asks a chief to fund a relationship the chief cannot sit in on, then asks a first responder to believe that promise when deciding what to say next.

Kutlerri.ai

A quiet catering phone is now part of a software supplier's pay problem. Kutlerri.ai says its fee comes from the business its agents generate or the costs they remove, putting the restaurant's next order inside the vendor's own economics.

Co-founder and CEO Kiran Rana and co-founder and COO Payal Tripathy have raised $4M from Houston-based Narahari Investments. The Texas company is expanding its Catering Sales Agent while developing more agents around restaurant revenue and operating costs. The financing was first reported September 30; Kutlerri distributed its Business Wire release October 15.

The National Restaurant Association says 42% of surveyed operators reported their restaurant was unprofitable in 2025. Catering gives that model somewhere concrete to start. An office lunch has a buyer to find, a menu to explain, a quote to approve and a follow-up somebody has to remember. A restaurant can have the kitchen and still lack the person who keeps those conversations moving.

Kutlerri runs the service for the operator. Its agents prospect for nearby businesses and handle incoming inquiries, proposals and follow-ups. The product also includes human sales support, and restaurant teams retain decisions such as quote approval. The labor behind an order remains part of the offering.

RASA was the development ground for 10 months. Kutlerri reports $450K in new catering revenue for the fast-casual group, more than 10 times its service fees. Those are company-reported customer sales, with food, delivery and other fulfillment costs still sitting inside the restaurant's profit calculation.

That puts attribution in the room early. Which orders did the service bring in? How are repeat customers counted? What happens when a booking changes? An outcome fee needs an agreed definition of the outcome, especially when the supplier keeps returning to the same local buyers.

The next products carry different jobs. Third-party sales, marketing-spend optimization and prime-cost optimization are in beta. Online sales and daily prep and waste control are listed as next. Each requires its own evidence of value; a catering result cannot do the measuring for food waste or labor spending.

Narahari Investments is backing a company taking responsibility for the commercial work around its software. As Kutlerri enters more restaurants, each menu, delivery radius and approval rule becomes part of the sales team's daily knowledge. The next office lunch has to arrive on terms the kitchen can live with.

Origin

The wall gets finished by a robot. Somebody on the crew still has to map the site and keep the material coming. At Origin, that division of work is where the construction AI story starts getting useful.

A contractor with more jobs than skilled people needs a way to accept the next project without exhausting the team already delivering the last one. Interior finishing puts that problem in plain view: compound, sanding, primer, awkward edges and the next trade waiting for a room.

Origin emerged from stealth on October 15 with $20M in total funding. Accel and Zacua Ventures co-led the financing. Unilazer Ventures, Schema Ventures, Z21, ISAI and Imad Ventures are also backers. The company has not broken the total into individual rounds.

Co-founder and CEO Yogesh Ghaturle previously built Perpule, acquired by Amazon in 2021. Co-founder Tushar Makkar leads autonomy after working on AI at Eightfold AI. Co-founder Ashish Daga leads hardware after precision automation work at Morphle Labs. Getting software, machinery and a construction crew to agree takes all three disciplines.

Their OG-1 robot applies compound, sands and primes walls. Origin says its Construction Action Model combines planning, an action policy and feedback control, trained with simulation and live job-site data. Interchangeable tools let the same hardware take on additional trades as the company develops them.

Origin describes a three-wheeled base that works up to 14 feet, with force control to maintain pressure while sanding. The work reaches beyond the easiest patch of wall.

That last part changes the buying conversation. A contractor evaluating a finishing robot is also deciding whether the supplier can become useful across more of the building. Each new task still has to fit the crew, the materials and the sequence of work around it.

The labor pressure is documented beyond the pitch. In AGC and NCCER's 2026 survey, 42% of respondents said shortages of their own workers or subcontractors' workers had delayed projects. A faster finishing workflow has value when it lets the rest of that schedule move.

Origin plans to use the funding for fleet buildout, deployments and larger pilots in the U.S. and Europe, with further pilots planned in Germany, Sweden, France and Saudi Arabia. Zacua founding partner Vivin Hegde has joined the board.

As OG-1 reaches more sites, the people mapping the next room and replenishing its materials become part of how Origin's intelligence earns a place in the working day.

Homa

Closing day gives a buyer several places to put the same dollar. Homa wants the fee for helping buy the house to leave more of that dollar available for financing it.

The Austin company announced a $5M Seed round on October 15, led by Quiet Capital with Restive Ventures and Cambrian Ventures participating. The money supports California expansion, mortgage capabilities inside the platform and a mobile app planned for later this year.

CEO and co-founder Arman Javaherian previously worked as a senior director at Zillow. CTO and co-founder Federico Campeotto brings a computer-science doctorate. Their product carries the home search into the work that follows: arranging a tour, understanding a property, making an offer and getting the transaction closed.

Homa puts licensed people around that work. Local specialists handle tours, licensed agents lead negotiations and coordinators manage the closing process. AI supports the research, including pricing and property analysis. The buyer gets different specialists without having to assemble the service from scratch.

The financial proposition has conditions worth reading. Homa charges 1% of the purchase price, with a $4,995 minimum. Where eligible seller-paid commission exceeds that fee, the remaining value can become a buyer credit of up to 2%, subject to the transaction and lender limits.

That gives representation a second job. Money returned at closing may help cover closing costs or fund a mortgage rate buydown. The credit becomes part of the financing conversation, where a household has to decide how much cash it needs now and what it can afford each month.

A lower quoted rate still comes with loan terms attached. The value of buying it down depends on the offer and how long the borrower keeps the loan. Homa's brokerage operates in California, Florida and Texas; its mortgage subsidiary currently discloses licenses in Florida and Texas. Those footprints deserve separate attention.

Quiet Capital partner Michael Bloch points to the coordination between brokerage and lending as part of the appeal. For the founders, that means taking responsibility for more of the moments when a home purchase can become confusing, expensive or slow.

California adds buyers who will meet the company through different specialists and still expect somebody to understand their whole purchase. A credit can bring them into the conversation. The agent, loan officer and closing coordinator have to keep the same conversation going until the keys change hands.

Sunbound / Richter Healthcare Consultants

An urgent move-in gives a senior living community very little time to become good at financial detective work. Sunbound is building around the decision made before anyone knows how long a resident's money will last.

That decision follows the operator through the stay. Somebody has to verify coverage, collect the family's payment, prepare the claim and deal with the payer that sends it back. The resident has already received the care while the money is still finding its way through the paperwork.

Co-founder and CEO Manny Cominsky and co-founder and CIO Stuart Mason have built Sunbound around those connected jobs. CTO Jiashu Wang leads the technology side of a business that pairs automation with people responsible for the complicated exceptions and keeping accounts current.

On October 15, Sunbound disclosed a $24M Series A led by TTV Capital, with Bling Capital, Fika Ventures, Live Oak Bank and Ziegler Linkage Fund participating. The company puts total funding at $36M and plans to use the capital for growth and product development.

Alongside the financing, Sunbound completed its acquisition of Richter Healthcare Consultants. The purchase price remains undisclosed. Richter brings more than 25 years serving post-acute and senior living operators, including the skilled nursing expertise Sunbound wants to expand around.

Jennifer Richter, CPA, the acquired firm's founder and CEO, stays with the business in a role focused on industry partnerships and product strategy. That puts someone who knows why a billing workflow gets complicated close to the decisions about what the software should do next.

Sunbound's admissions tools screen financial readiness before move-in. Private Payments handles digital payments and family follow-up. Its revenue cycle offering works through eligibility, claims, denials, appeals and reconciliation above the operator's existing EHR.

The combined companies say they are on track for $3B in annualized payment volume. That is a processing projection, with a lot of individual residents and disputed payments underneath it. Sunbound's job grows with every piece of the financial relationship it takes on.

For an operator, putting these services with one supplier could mean fewer handoffs between the people checking coverage and the people explaining the unpaid balance. Jennifer Richter's team now helps carry those handoffs into a larger business, while the next family is already trying to arrange a move-in.

Chariot Claims

The individual arbitration clause has acquired an operations department. Chariot Claims is building the machinery for thousands of consumers to pursue separate claims through partner law firms, even when a company's contract closes the class-action route.

Founder and CEO Zim Hang saw the opportunity in the work those firms struggled to absorb: finding people, checking eligibility, collecting evidence and keeping each claim organized. CTO Jon Chen and COO Sharon Cao are part of the current leadership team taking that work into software and daily operations.

Chariot announced a $10M Series A on October 15, led by Silverton Partners and general partner Mike Dodd. Indicator Ventures, Lightbank, Neon Fund, ZVC, Rosecliff Venture Partners, Alumni Ventures, FJ Labs, Graph Ventures and Vermilion Ventures also participated. Total funding reaches $15.7M, including the $3.6M Seed round reported for April 2025.

The business has two audiences to satisfy. Consumers want a manageable way to pursue alleged privacy violations, hidden fees and difficult-to-cancel subscriptions. Law firms need people whose circumstances and supporting documents fit the case. An enthusiastic click still needs to become something an attorney can use.

Chariot says it covers claimant-acquisition costs upfront, collects evidence and manages payouts. Zim Hang told Crunchbase News that the company charges firms flat service fees; the firms handle the legal claims on contingency. Chariot itself is not a law firm. That division gives the software company a very concrete job between the advertisement and the attorney.

The company reports processing more than 200,000 pre-qualified claims a month. That is an operating-volume claim, with technology and human review behind it. A submitted claim, a qualified participant and a successful recovery occupy different points in a process that can take 18 months or longer, according to Chariot's own site.

The financing will support AI development, hiring and operations. It also gives Zim Hang's team more capacity to stay involved as the companies named in potential cases object to its outreach. Chariot has described cease-and-desist letters concerning its advertising, making the cost of reaching consumers part of the business it has to manage.

A consumer's receipt may look trivial on its own. Chariot is organizing the people, documents and follow-through that give a law firm somewhere to begin with thousands of them.

Opsin

Giving an AI agent permission can leave its owner with a second job: explaining why it needed that permission. Opsin is building around the distance between an approved connection and an appropriate action.

The company announced $21M in funding on October 15, led by SYN Ventures and Race Capital. Co-founders James Pham, CEO, Oz Wasserman, CPO, and Jeremy Mailen, CTO, are expanding a business that began with enterprise data oversharing into security for agents with more authority to act.

Pham and Wasserman met at Abnormal Security. Mailen brought engineering experience from Trifacta and Divebell, giving the team a useful combination for a problem that crosses machine learning, security and enterprise data.

An agent can use its owner's credentials to reach information the person asking the question cannot access directly. A connection may be technically allowed while its consequences take everybody by surprise. The security team inherits the explanation after the employee has already found the shortcut. For the buyer, the next approval has to survive a conversation with the people whose data the agent can reach.

Opsin connects an agent's purpose, owner, permissions, data, tools and activity through its Risk Graph. Its common model spans Microsoft 365 Copilot, Copilot Studio, Claude Enterprise, ChatGPT Enterprise and Gemini, so the buyer can assess agents across the systems employees actually use.

That context has a practical job. Opsin says it can identify risky departures, quarantine an agent, block an action or notify the owner about a fix. Its Agent Intent feature also examines an agent's configuration before a task runs, giving security a chance to question excessive access before somebody puts it to work.

Opsin reports nearly 20 enterprise customers, including Culligan, Barry-Wehmiller, Wellstar Health System and UiPath. The new financing supports product development, hiring and broader support for regulated enterprises. Its earlier $7M Seed was announced in April 2025; the latest disclosure leaves the new round's stage and lifetime funding total unspecified.

Selling into that environment means helping security make a decision it can explain to the business. Blocking a useful workflow has a cost, and allowing the wrong one can reach far beyond the person who created it. Opsin's founders now have more funding to carry that judgment across more agents, more connectors and the next department asking to turn its automation on.