CarbonAI Announces $12M for Commercial Real Estate Lending

After selling Nativ to Built, Jeff Saul and Josh Kaufman returned to commercial real estate software with a different job for the machine. At CarbonAI, the co-founders are building agents to prepare the analytical work around a loan while the lender keeps its own models, reporting conventions and authority over decisions.

The New York company announced $12M in funding on October 15, 2026, led by Cowboy Ventures with Timber Grove Ventures, Bling Capital and Max Ventures participating. The financing puts capital behind a particular approach to financial-services AI: enter the work the institution already understands, then carry more of it through to a reviewable result. For commercial real estate lenders balancing new transactions with existing portfolios, that is a proposition about operating capacity as much as software.

What CarbonAI's funding announcement establishes

CarbonAI's announcement identifies the amount and investors without assigning the financing a named Series or publishing a valuation. The careful description is $12M in announced funding; the available disclosures do not justify adding that amount to an assumed previous round. Nor does an investor's early-stage focus establish the precise structure of a portfolio company's transaction.

CEO Jeff Saul and COO Josh Kaufman bring experience in commercial real estate finance and a previous lender-software business. CarbonAI's current team page describes a path through firms including Starwood, Morgan Stanley, Argentic and Guggenheim before Nativ and Built. That history is commercially relevant because the new product asks customers to hand over work shaped by their own lending standards. Familiarity with those standards gives the founders a way into a conversation that a polished demonstration alone may struggle to sustain.

Cowboy Ventures led the financing. Partner Caroline Duffy describes the founders' industry experience and existing lender relationships as reasons for backing the company. That is an investor's account of its conviction, rather than an independent product benchmark. It nevertheless explains the fit: the financing backs people who have already spent time selling software into the profession whose work they now want agents to perform.

How the agents enter a lender's workflow

CarbonAI's implementation description begins with the customer's documents and templates. Rent rolls, leases, operating statements and loan documents supply the inputs. The software extracts and reconciles information, maps the lender's existing models and prepares an underwriting workbook with an accompanying investment memo. Work requiring judgment is routed to the lending team.

Keeping the workbook matters because a model is also a record of institutional choices. It reflects the questions a lender asks, the assumptions it accepts and the way a reviewer expects to inspect a transaction. A supplier that can work within those conventions may reduce the amount of organizational change attached to a purchase. The implementation still requires careful review, but the customer has a familiar place to begin that review.

The company also describes recorded agent activity, source-linked figures and rule-based reconciliation. Those mechanisms address different questions: where an input came from, what the software did with it and whether the resulting arithmetic agrees. A credit professional can use those records to investigate an output rather than treating a finished memo as an answer whose preparation has disappeared.

CarbonAI states that its platform is SOC 2 Type II compliant, hosted in the U.S. and designed to keep customer data out of model training. These are company representations; an institution considering deployment would need to examine the relevant controls and documentation itself. The combination of financial analysis and sensitive transaction information makes that work part of the purchase.

Why the loan book keeps the work coming

Commercial real estate lending carries a calendar that software vendors cannot wish away. The Mortgage Bankers Association's February 2026 maturity survey identified $875B of commercial mortgages scheduled to mature during 2026, or 17% of outstanding balances in the surveyed market. That was down 9% from the amount scheduled for 2025. The figures describe elevated refinancing activity without supporting a claim that the maturity wave is still growing.

For an analytical-software business, the useful distinction is between winning a new loan and continuing to understand a loan already held. CarbonAI includes quarterly reporting and asset surveillance in its product scope. Updated operating information has to be reconciled with earlier reports, while lending teams continue to assess new opportunities. The same institutional standards can therefore matter repeatedly across the life of a relationship.

This creates a commercial route beyond a single document-processing task. CarbonAI describes separate uses for origination, underwriting, credit and asset management. Connecting those roles through traceable analytical work could make the product more valuable to a customer already using it in one part of the loan process. That is an interpretation of the product design, with the extent of actual adoption across each customer's teams still undisclosed.

What the early customer evidence can tell us

CarbonAI reports that early customers saved as much as 90% of the time on selected underwriting tasks, and that contractual revenue reached seven figures within six months of its early-2025 product launch. The scope of both claims matters. A task-level saving cannot be applied automatically to an entire loan process, and contractual revenue does not establish recognized sales or annual recurring revenue.

Greystone provides a named customer perspective through the testimonial on CarbonAI's website. The account emphasizes the team's willingness to understand due diligence collection and build around the lender's actual process. It gives the company's positioning a practical example: listening closely can determine which work becomes useful to automate before a customer expands its use of the platform.

CarbonAI has not disclosed a detailed allocation of the funding or a comprehensive set of independently verified operating outcomes. The work now unfolding is specific enough without inventing either. Jeff Saul and Josh Kaufman are bringing agents back into a profession they already served, carrying each lender's familiar models from the next transaction into the reviews that follow it onto the books.