PARES AI Is Building the Bloomberg Terminal for Commercial Real Estate Brokers

The Macro: CRE Brokers Are Drowning in Manual Work

Commercial real estate is a strange industry. The deals are enormous. A single transaction can involve tens of millions of dollars. The brokers facilitating those deals earn substantial commissions. And yet the technology stack most of them use looks like it was designed in 2008 and never updated.

I am not exaggerating. Talk to any CRE broker and ask them about their workflow. They will describe some combination of Excel spreadsheets for deal tracking, a generic CRM that was not built for real estate, manual comp research using CoStar or LoopNet, and marketing materials assembled in PowerPoint or Canva. The underwriting process involves pulling data from five different sources, dropping it into a model someone built three years ago, and hoping the formulas still work.

CoStar dominates the data side of commercial real estate with a near-monopoly on property listings and comparable sales data. Crexi has made inroads on the marketplace side. Reonomy tried to build a data intelligence layer and got acquired. Buildout handles marketing materials for some brokers. But nobody has put together a single platform that handles the entire broker workflow from prospecting to closing.

This is a pattern I have seen before in other professional services. Lawyers had fragmented tools until Clio consolidated the stack. Accountants had the same problem until Karbon and similar platforms showed up. CRE brokers are next. The question is who gets there first with a product that actually works.

The AI angle matters here because CRE underwriting is fundamentally a data processing problem. Rent comps, operating expense analysis, cap rate calculations, cash flow projections. These are tasks where AI can genuinely save hours of analyst time per deal. The brokers who can underwrite faster can bid on more deals and win more listings.

The Micro: One Founder Going After a Massive Vertical

PARES AI is building an all-in-one platform for CRE brokers that combines CRM, skip-tracing, AI underwriting, and automated marketing document creation. The pitch is simple: stop switching between six different tools and do everything in one place.

Zihao Wang is the co-founder and CEO, running a three-person team out of San Francisco as part of Y Combinator’s Summer 2025 batch. The product is live with a functional sign-in and demo scheduling through Calendly.

The platform breaks down into three main modules. AI Underwriting handles rent and sale comparables, automated rent roll parsing, T12 financial statement coding, and pulls from both public and proprietary data sources. AI Marketing generates professional presentation materials branded with the broker’s company colors and logos. AI Asset Management tracks budget variances, benchmarks NOI, DSCR, returns, and waterfall distributions, and produces professional reporting.

What caught my attention is the skip-tracing integration. For brokers doing cold outreach to property owners, skip-tracing (finding contact information for property owners from public records) is a critical workflow that usually requires a separate subscription to a service like Reonomy or BatchLeads. Baking it into the same platform where you manage your pipeline is a smart consolidation play.

The technical stack is built on Next.js with Clerk for authentication, which tells me this is a modern web app rather than a legacy enterprise deployment. That matters because CRE technology has historically been sold through painful enterprise sales cycles with on-premise installations. A modern web app that brokers can start using immediately changes the adoption curve.

The Verdict

I think PARES AI is attacking the right market at the right time. CRE brokers are underserved by technology and the ones who adopt AI-powered underwriting tools will have a genuine competitive advantage in deal velocity. The question is whether a three-person startup can build enough depth in each module to replace the specialized tools brokers currently use.

The risk is the classic all-in-one trap. When you try to replace six tools with one platform, you need to be at least 70% as good as each specialized tool on day one. If the CRM is worse than Salesforce, the underwriting is worse than Argus, and the marketing is worse than Buildout, brokers will stick with their fragmented stack even though switching between tools is painful.

Thirty days from now, I want to see how many brokers have moved their primary workflow into PARES versus using it as a supplement. Sixty days, the question is whether the AI underwriting is producing output that brokers actually trust for client-facing work. Ninety days, I want to know the average deal size flowing through the platform. If it is small deals only, they have a positioning problem. If large institutional deals are showing up, they have something real.