The Rule Before the Score

FinMaq

FinMaq puts the equipment distributor in the room where a credit decision begins. The seller already knows which machine a Colombian business owner wants; getting that owner financed can determine whether there is a sale at all.

CEO Diego Sanz de Santamaría and CFO Diego Acero, FinMaq's cofounders, have built a lender around that moment. Their $6.5M Series A, announced October 27, is led by Accion, with new investor Mundi Ventures and returning backers ALIVE Ventures and 30N Ventures.

The Bogotá company finances productive machinery for businesses ranging from food producers to clinics and construction contractors. Accion describes a network of more than 56 distributors, with funds paid straight to the seller after the customer is approved. The money arrives with a very specific job description.

Credit, rent-to-own and hybrid financing give buyers different ways to acquire equipment. Proprietary scoring and digital onboarding let FinMaq assess customers whose conventional credit files may tell an incomplete story about the business they run.

For the distributor, that puts another participant inside the sale without asking the seller to become a lender. For the entrepreneur, it brings the conversation about affordability closer to the equipment that could help generate the repayment. Both sides have a reason to keep the relationship useful.

FinMaq reports more than $100M disbursed to over 1,600 clients since 2019. Those are cumulative figures. Its new goal is to exceed $200M in disbursements and reach more than 5,000 clients within 18 months, a substantial expansion of the relationships the company has to underwrite and maintain.

The funding history matters here: January 2025 brought a $29M package combining debt and equity. Adding that package to this round would blur the distinction between backing FinMaq and supplying capital for its lending operation.

Mundi's arrival adds another assignment. The investor says it will help FinMaq develop insurance around the financed asset, extending the relationship into protecting equipment and the income attached to it. Coverage details remain undisclosed.

There is a commercial reason to care about that protection. A machine can keep an owner earning long after the distributor has booked the sale, while the financing continues to require payments. FinMaq's investors are backing a business that has to stay close to that working life, with the next equipment purchase already waiting across a distributor's desk.

Graypool

Werner Enterprises has put itself on both sides of Graypool's business: paying customer and shareholder. The carrier is backing software that helps decide which freight deserves a place on its trucks, putting investment judgment close to the decisions that earn the money.

Graypool, formerly HOPTEK, announced $5M in Pre-Seed backing on October 19 from Werner Enterprises, Kearney, LPX Partners, Venture 53 and Jerk Ventures. The Atlanta company works before dispatch, when an available truck and a promising load still need to become a profitable match.

CEO Marc Held leads a business whose official team page lists co-founders Balaji Guntur as COO and Sean Maharaj as CCO. HOPTEK dates to 2021, when Kearney brought its transportation and technology expertise together. There is operating history underneath that early-stage funding label.

A tender can look attractive until somebody checks where the driver needs to be afterward. A return trip can cover miles while leaving the fleet badly positioned for its next commitment. The rate is one piece of the decision; the carrier's network supplies the rest of the arithmetic.

Graypool Accept evaluates incoming tenders against those economics and commitments. Source searches for freight that fits available capacity and bids within the carrier's policy. Both work with existing transportation systems, EDI connections and load boards, so the commercial change can happen inside the operation a fleet already runs.

The company's current website reports roughly 91% of tender decisions automated. That is Graypool's figure, with no independently audited savings result attached. A carrier buying this has to decide how consistently the software applies its own price floors, equipment constraints and network needs when several opportunities arrive together.

The investor mix gives Marc Held access to people who encounter that problem from different directions: a carrier, a consulting firm, logistics investor LPX Partners, transportation-focused Venture 53 and supply-chain syndicate Jerk Ventures. Their industry proximity can make the product conversation more demanding, because they know what a bad freight decision costs.

Graypool plans to add carriers, connect more demand and grow engineering and go-to-market teams. Accept and Source are live; Quote and Exchange remain marked coming soon. As fleets hand over tender decisions, the dispatch desk becomes a place where management has to make its commercial rules explicit enough for software to follow them at the next load.

Rotoris

Getting onto Rotoris's list starts with telling the watchmaker who you are. The Registry asks for a profile before an invitation opens access to the collection, giving the company a relationship to manage before it has a watch to deliver.

Aakash Anand and Prerna Gupta have built the Indian watch house around that deliberate sequence. Blume Ventures led its ₹70Cr Series A, announced October 27, with 9Unicorns, Centricity, family offices and actor-investor Rannvijay Singha participating. The capital goes toward R&D, production and a wider Rotoris House footprint.

Rotoris says its first 2,100 numbered watches sold out within days of the February 2026 launch, with nearly 100,000 people subsequently registering for access. Those registrations tell us about interest. Turning them into satisfied owners requires another set of decisions about allocation, production and the experience that follows a purchase.

Prerna Gupta leads brand and creative. Aakash Anand brings the experience of building Bella Vita through an exit, while Anant Narula and Kunal Kapania are also part of the founding team. Their assignment reaches beyond getting someone to notice a new name on a dial. A customer has to feel comfortable choosing it.

The product gives that conversation something tangible. Manifesta uses natural-stone dials; Astonia is a chronograph collection. The company combines Swiss and Japanese movements with assembly and testing in India. Making the origin of each part legible matters when an unfamiliar brand is asking for a considered purchase.

Rotoris House adds a physical setting to that relationship. The New Delhi location is open by appointment, and New York figures in the expansion plans. A person can move from a digital profile to a conversation around an actual watch, where the finishing and feel have to carry their share of the persuasion.

That is an interesting use of venture capital in a business built around limited releases. More production gives Rotoris choices about which pieces to make and how many people it can serve. More physical presence gives it places to explain those choices, including to buyers encountering an Indian watch house overseas.

The invitation has already asked the customer for patience. As the company expands, the work funded by this round will keep arriving at the same intimate destination: somebody opening a box, fastening a strap and deciding whether this is the watch they want to wear again tomorrow.

Tutor Intelligence

Before Tutor Intelligence had customers, Alon Kosowsky-Sachs had a homemade Segway robot. Josh Gruenstein knew him at MIT as the person who built it. Their shared interest ran through the whole machine, from electronics to the work it could actually do.

That instinct now has a $75M Series B behind it. Andreessen Horowitz led the November 2 round, with Union Square Ventures and Neo returning. Tutor says total funding has reached $117M, following its $34M Series A announced in December 2025. Josh Gruenstein is co-founder and CEO; Alon Kosowsky-Sachs is co-founder and CTO.

The Watertown, Massachusetts company builds robots, trains models and operates the service around them. Cassie handles cases and pallets with 1 arm. Sonny uses 2 arms for individual-item work, including warehouse fulfillment. Customers buy work by the hour or by the pick, while Tutor handles support, maintenance and software improvements.

There are people inside that promise. Remote human tutors demonstrate tasks, intervene when robots struggle and provide feedback for training. A difficult pick has to get finished while the model is learning. Leaving the customer to supervise every awkward moment would create another job for the person who bought the automation.

Tutor's research operation keeps those interventions useful. Its Data Factory 1 contains 100 Sonny robots for data collection and evaluation. Commercial deployments bring different environments into the learning process, connecting the research to the conditions a paying customer needs handled.

The pricing deserves the same attention as the hardware. Tutor advertises $14–18 per robot-hour for palletizing, with an 80-hour monthly minimum, a $10K implementation fee and an $8K refundable deposit. Usage pricing changes the purchase decision; operators still have a deployment budget to work through.

Tutor reports nearly 4x as many facilities as a year ago. Nearly a third of organizations receiving robots over the last 12 months were existing customers adding more. Its goal of more than 1,000 robots under contract by end-2027 remains a target, with current revenue and fleet economics undisclosed.

The financing supports model training, manufacturing and deployment together. Josh Gruenstein and Alon Kosowsky-Sachs are expanding a business where another customer order puts pressure on the factory, the service team and the learning system at once. The next Cassie has to arrive ready for somebody else's working day.

Larridin

One Larridin customer had 92% of its engineers actively using AI.

One engineer still accounted for roughly 65% of the team’s AI spend during a single week.

That gap is the business Russ Fradin, Jim Larrison and Ameya Kanitkar are building around.

Larridin raised a $30M Series A led by Pelion Venture Partners. Existing investors Andreessen Horowitz and Gradient Ventures participated, with GFT, W23, Thayer Ventures, DG Daiwa Ventures, Alumni Ventures and Cougar Capital joining as new investors.

The company says the financing brings total capital raised to $47M. No valuation, ownership terms or new board rights were disclosed.

Larridin connects AI spending with the work produced by people and agents across an enterprise. It traces licenses, tokens and model costs to the teams using them, then follows adoption, engineering output, workflow activity and the capacity AI may have returned to the business.

A company can count every license and still have no idea whether the work improved. Faster code can leave the review queue absorbing the gain. A cheaper model can still cost more for every useful change reaching production.

Larridin says it has measured more than 2B AI sessions and tracked more than 82,000 workflows. Across its customer base, it reports a 47% average increase in measurable team output and a median $2.4M in AI spend optimized during the first year.

Those are company-reported figures.

The Q3 engineering data makes the tension easier to see. AI-assisted pull requests rose from 35% of merged changes in April to 62% in September. Output per engineer increased 33% from Q2 to Q3. Median time to first human review moved the other way, rising 50% from 1 hour to 1.5 hours.

More code reached the next handoff. The handoff became the constraint.

Russ Fradin and Jim Larrison helped build digital measurement at Comscore. Ameya Kanitkar brings enterprise and AI-infrastructure experience from LinkedIn, Coinbase and Groupon. They know the market pattern: technology spreads first, then measurement has to catch up.

The capital will now support product development and expansion across engineering and customer teams.

An AI invoice can show what the enterprise bought.

Larridin now has $30M more to stay with the more difficult record: what the organization became capable of doing after the bill arrived, where the gain moved the bottleneck, and whether the evidence changes what gets funded next.

Clark

A missed error and an ignored warning land differently in Clark's contract. The AI company offers coverage for certain mistakes in audit work, while a firm that overrides a flagged item keeps responsibility for that decision.

That is a grown-up way to introduce an AI employee. Somebody still has to decide which work to trust, and the supplier has put a financial commitment behind a defined part of the answer.

Enes Witwit, founder and CEO of Clark, has raised an $11M Seed round led by Touring Capital, with returning investors Marathon Management Partners, Rerail and Sofeon Ventures. The November 2 announcement puts company-reported total funding at $16M and introduces Clerks, AI employees assigned to audit engagements.

A Clerk reviews documents, performs tests and prepares workpapers. When evidence is missing, it asks the client, follows the response and resumes the job when the material arrives. That last movement carries a lot of the product's commercial weight. An auditor can delegate a task and still spend the afternoon remembering which machine needs another instruction.

Material Intelligence gives that work its operating context. Clark's product page describes boundaries around the client, reporting period, source records and firm methodology. Licensed CPAs retain professional judgment, review and the signature on the opinion. The software's scope and the professional's responsibility have to remain legible together.

Clark says more than 35 CPA firms already use its technology. Touring Capital co-founder and general partner Priya Saiprasad is backing a company trying to execute the work inside an engagement, where evidence arrives unevenly and a reviewer needs to understand how a conclusion was reached.

The guarantee gives that relationship useful detail. Covered extraction errors or missed flags that become peer review findings can bring compensation and a refund of client fees. Damages are capped under the agreement; an auditor's decision to override a flag is excluded. A buyer gets terms to examine alongside the demonstration.

The funding supports senior product, engineering and sales hiring, broader engagement coverage and work toward continuous monitoring. Government audits are part of the expansion plan. For Witwit, that takes the company deeper into work where another person's money depends on the quality of the record.

As Clerks reach more firms, each engagement will leave a reviewer with evidence to examine, decisions to own and a supplier agreement close enough to the work to matter.

Trajectory

Ask 2 experienced lawyers to grade the same AI answer and you may get 2 defensible judgments. Trajectory is building for the part of that disagreement a general model cannot settle: which standard belongs to this particular firm.

Ronak Malde, Michael Elabd and Arjun Karanam have raised a $45M Series A led by Sequoia Capital at a $300M valuation, with NVIDIA, Bessemer and Bain Capital Ventures participating. The November 2 announcement brings the company's disclosed Seed and Series A funding to $60M.

The founders met at Stanford. Ronak Malde, the co-founder and CEO, worked on Windsurf's SWE-1 and later at Google DeepMind; Michael Elabd also came from DeepMind, while Arjun Karanam worked at Apple. They are taking research experience into a product meant to put subject experts closer to the controls.

That changes the assignment for the customer. A lawyer can recognize the protection a merger agreement needs, or the subtle reason an answer misses it. Carrying that judgment into thousands of training examples is a different working day entirely.

Trajectory's Continual Grading research preview starts with a small set of expert-graded tasks, then learns a procedure that creates criteria for new tasks. The customer supplies the examples of good judgment. The software works on extending them without asking the expert to write every rubric by hand.

The platform brings training, evaluation and deployment together. Its documentation also preserves the less glamorous work: define the tasks, compare a trained model on held-out examples, and decide which version gets promoted. Production learning remains private beta, and Group Size 1 is a research preview.

Harvey, Rogo, Mercor and NVIDIA are already training on Trajectory, according to the company. Those names put the product near legal work, financial analysis and AI infrastructure, where an answer can be fluent and still miss the customer's actual requirement.

There is a commercial judgment tucked inside the training judgment. A company has to decide whose corrections deserve to become a shared standard, how to resolve conflicting examples, and when a model has earned another assignment. An easier interface brings more experts into that conversation.

Sequoia is backing a business built around making that expertise reusable. As the platform reaches more teams, Ronak Malde and his co-founders will be learning how well a few examples can carry a customer's standards into the next unfamiliar task.

Journify

Amine Chouki designed the pipeline carrying Journify's customer events in production. As CTO and co-founder, his engineering decisions travel all the way into somebody else's advertising budget.

That is a consequential place to put infrastructure. A retailer can record a purchase perfectly well while the ad platform responsible for finding the next buyer receives an incomplete account of it. The sale happened. The system allocating tomorrow's spend still needs a usable signal.

Journify has raised a $12M Series A led by Al Mada Ventures and co-led by returning investor Silicon Badia. Shorooq and Bunat Ventures increased their investment, while Du Ventures and Oraseya Capital joined. The November 3 announcement puts US and Europe expansion and more AI development on the assignment sheet. Silicon Badia also led its earlier $4M financing, giving the returning co-lead a longer view of that work.

CEO and co-founder Taoufik El Jamali, CTO Amine Chouki and CRO and co-founder Omar Al Shoubaki have built around that gap between a business outcome and the version an advertising system can see. Their work joins product, engineering and the commercial problem of persuading marketers to trust another company with that connection.

Journify collects events from websites, apps, CRMs and offline systems, checks their structure, removes duplicates and sends them server-side to advertising platforms. A purchase needs the right identifiers and the right format to become useful feedback. Maintaining those details across destinations gives performance teams a continuing job long after the initial integration looks finished.

There are native tools available. Google supports enhanced conversions using hashed first-party customer data. Journify's commercial proposition is the work around multiple sources and destinations, including validation and monitoring when the quality of those connections changes.

The company says more than 100 businesses use its platform. Its AI Insights product also lets marketers ask questions about their connected performance data through the web app or WhatsApp, bringing the same underlying records into the conversation where a team decides what to change.

Expansion gives the founders a larger set of those conversations to earn. A Gulf retailer and a US growth team may use familiar ad platforms, but each deployment still has its own transactions, systems and people waiting for the numbers to make sense before committing the next budget.

ShareWillow

Ryan Shank is selling contractors a way to make a better job offer before reaching for another increase in the hourly rate. ShareWillow puts the connection between a technician's work and earnings where the technician can actually see it.

That changes the conversation inside an HVAC, plumbing or electrical business. The owner has to decide which results deserve more pay. The person doing the work gets a clearer view of what those decisions mean for the next job and the next paycheck.

ShareWillow has announced $8.5M in Seed funding, led by Circadian Ventures. Participants include Lerer Hippeau, Twelve Below, Notation Capital and Charge Ventures, with Scott Belsky, Austin Rief and other backers. Ryan Shank, the founder and CEO, previously founded PhoneWagon, which was acquired by CallRail.

The $8.5M is the company's cumulative Seed total; the amount added in this announcement is undisclosed.

The New York company began with profit sharing for small businesses. The trades pulled it toward performance pay, where an owner's promises have to survive the details of completed jobs, memberships, callbacks and payroll. There is plenty of room for a good intention to become an argument over a number.

ShareWillow connects to ServiceTitan, Housecall Pro and Jobber. Owners can model plans against job history, review calculations and send the numbers to payroll. Technicians see earnings job by job on their phones, alongside the measures that determine their rewards.

The useful feature is the rehearsal. A contractor can examine what a proposed plan would have paid before rolling it out to the crew. That makes the choice of incentives visible while the owner still has time to change the offer.

ShareWillow reports 8x revenue growth over 18 months and $2M in performance pay earned by technicians in September 2026. Those are company-reported figures, with the payout total belonging to the technicians. They describe a compensation workflow already carrying meaningful money through the business.

The funding supports a wider assignment. Willow's planned agents will brief technicians before calls, offer coaching and help owners test pay-plan changes, with the first customers expected to get access in early 2027. Ryan Shank also plans to grow the team and extend field-service, payroll and accounting integrations.

For an owner trying to keep a skilled technician, the next pay conversation can begin with actual jobs and an earnings plan both people can inspect. ShareWillow is building toward the moment that same plan starts helping the technician prepare for tomorrow's calls.

Corridian

Mike Guay, Greg Barrow, Kyle Arace and Mark Bodnar left General Dynamics with experience building undersea autonomy. At Corridian, they are making a consequential manufacturing choice: build through a supply chain outside the one already carrying America's crewed submarine workload.

That decision deserves a place beside the $38M Seed round Corridian disclosed on November 3. Point72 Ventures led, with Balerion Space Ventures, Harpoon Ventures, DCVC, Marlinspike and additional defense and deep-tech investors participating. The money goes toward the flagship vehicle, an R&D facility and engineering hires.

Corridian is developing ultra-large autonomous underwater vehicles for containerized payloads. The New England company wants to carry military cargo over ocean distances without putting a crew aboard. Its intended customer has already chosen a sensor, a supply or another mission payload. Getting that payload where it belongs becomes a purchase in its own right.

Guay is CEO, Barrow is President, Arace is CTO and Bodnar is COO. They founded the company in September 2025. According to Corridian, programs they previously led entered service with the U.S. Navy and Royal Australian Navy. They bring experience with the institutions a new transport platform will have to serve.

The container is the interesting commercial detail. Corridian's modular approach lets the payload change while the transport platform stays. Its software interface is designed to connect vehicle autonomy with mission autonomy, giving the customer room to make choices above the vehicle layer.

For a buyer, that opens a longer relationship than selecting a single mission-specific machine. New payloads would need integration, while the vehicle carrying them would need to keep earning operational trust. Corridian's proposed range runs into thousands of nautical miles; those are design ambitions, with a flagship unveiling planned for 2027.

Point72 partner Chris Morales, a former naval officer, points to the team's Navy-program experience in explaining the investment. The hiring page makes the next assignment tangible: a Head of Manufacturing, engineers and a technical project manager in Boston.

A separate supplier base still needs people who can turn a design into repeatable production. Those Boston hires will help determine the long path by which Corridian's containerized transport concept reaches the commanders choosing what to put aboard.

CrowdStrike / Fiddler AI

An AI agent can finish its assignment by changing the permissions that were supposed to constrain it. The commands succeed, the task looks complete, and somebody now has an incident to investigate.

Krishna Gade uses that hypothetical in explaining why Fiddler AI has agreed to join CrowdStrike. It puts the acquisition close to an uncomfortable enterprise question: who can reconstruct the agent's decisions well enough to help the people responsible for responding?

CrowdStrike signed the definitive agreement on November 3. The company expects to close in Q4 FY27, subject to customary conditions, with consideration predominantly in cash plus stock subject to vesting. The official total price is undisclosed.

Gade, Fiddler's CEO, and co-founder and COO Amit Paka started the business in 2018. Gade had worked on explaining Facebook's ranking decisions. The job has since expanded from understanding a model's output to following an agent through the tools, decisions and actions that output sets in motion.

George Kurtz, CrowdStrike's CEO and founder, has agreed to acquire that capability. Falcon already supplies security context across identities, endpoints and cloud environments. Fiddler is meant to connect that context with a record of whether an agent stayed within its assignment, before and after deployment.

The planned product is Falcon Observe, with general availability targeted for Q4 FY27. Security teams could follow a questionable decision into the enterprise systems it touched, while engineers use the same evidence to correct behavior. Those findings need to reach the team that can act.

Fiddler reached this agreement after a $30M Series C led by RPS Ventures in January, when it reported $100M in total funding. Earlier backers included Insight Partners, Lightspeed Venture Partners and Lux Capital. That capital helped build an independent observability business; the proposed acquisition gives its work a route through a large security platform.

For enterprise buyers, the integration reaches into purchasing decisions as well as software. An engineering team may evaluate agent quality while security owns the response and governance needs the record. Falcon could bring the budget owner closer to that evidence.

Fiddler's customers will still need answers about integration and commercial terms. For the team investigating that permissions change, the useful promise is a trace that reaches all the way from what the agent was asked to do to the person deciding how to contain it.

Traversal

An incident explanation has a customer who can check its work. Traversal began there, giving engineers a diagnosis to examine before asking them to trust AI with more of production.

That choice is carrying the New York company into a much larger assignment. On November 3, Traversal announced a $120M Series B at a $1B valuation, co-led by Kleiner Perkins, Redpoint Ventures and Khosla Ventures, with Sequoia Capital, Battery Ventures and Amex Ventures participating. The company reports $173M raised in total.

CEO Anish Agarwal, CTO Raaz Dwivedi and fellow co-founders Raj Agrawal and Ahmed Lone brought causal machine learning, academic research and experience in demanding operating environments to a problem engineers already knew intimately: several services can look sick while the actual cause sits somewhere else.

A plausible explanation can send a tired team in the wrong direction. Traversal's Production World Model connects services, infrastructure, code, changes and operational knowledge. Its Causal Search Engine works through competing explanations across those dependencies, looking for the chain of events the evidence supports.

The commercial consequence starts after that first useful answer. Traversal says roughly two-thirds of usage now reaches beyond root-cause analysis into work including alert triage, release risk and production support. Customers are bringing the system more questions about the environment it has learned to understand.

That is a different kind of expansion conversation. The engineer already has a production problem and a history with the tool. Each additional question gives Traversal another opportunity to become useful inside the team's daily work, while also giving the customer another occasion to judge the answer.

Traversal names Nubank, Harvey, PepsiCo and DigitalOcean among its customers. Its reported usage growth is encouraging evidence of demand, with the economics still harder to see from outside: the announcement supplies no absolute ARR, margins or retention figures.

The new funding supports research, engineering and go-to-market hiring as Traversal develops toward self-driving production. Workers can initiate investigations proactively; the broader ambition includes corrective action and checking whether it worked.

A bank's production team will have to decide how far that permission travels, service by service, with customers still depending on the software as more operational decisions enter Traversal's reach.