Analysis
CarbonAI, a New York-based startup, raised $12 million led by Cowboy Ventures, with Timber Grove Ventures, Bling Capital and Max Ventures participating, Pulse 2.0 reported. The company builds AI agents that handle analysis at every stage of a commercial real estate loan: underwriting, due diligence, investment memos and ongoing post-close monitoring, with every output tied back to its source document and anything requiring human judgment routed to lending teams rather than auto-approved.
Co-founders Jeff Saul (CEO) and Josh Kaufman both worked in CRE lending at Starwood Capital, Morgan Stanley and Guggenheim before this, and they've built in the space before: their earlier company, Nativ, was a deal-management platform that construction-tech firm Built later acquired. That history is the core of CarbonAI's pitch, founders who understand loan committees from the inside building the automation layer for the same workflow, rather than a generic AI-agent startup bolting on a CRE use case.
CRE lending technology already has incumbents working adjacent angles: Cherre and Dealpath sell data and deal-management platforms to real estate investors, Northspyre focuses on development and construction cost tracking, and traditional loan origination software from Jack Henry and Finastra still runs much of the back office at regional and community banks that hold the bulk of CRE loans. CarbonAI's bet is that none of those platforms fully automates the underwriting judgment calls, document review and ongoing monitoring that currently consume analyst hours at lenders and debt funds.
The round is small relative to the AI-agent funding rounds making headlines elsewhere this month, and neither CarbonAI's valuation nor its revenue were disclosed, so there's no way yet to judge whether lenders are actually paying for this or still piloting it. CRE lending is also a sector still digesting elevated interest rates and regional-bank stress from recent years, a tougher environment to sell new software into than it would be in a lending upcycle, even with a strong founding team.