Closera Thinks AI Can Save Commercial Brokers 35 Hours a Week

The Macro: Commercial Real Estate Runs on Manual Labor and Nobody Talks About It

Commercial real estate is one of the last major industries where deal-making still runs on brute-force human labor for tasks that have no business being manual. A broker working on a sale needs to produce an offering memorandum. That document includes property photos, financial projections, market comparables, tenant information, lease abstracts, demographic data, and a narrative pitch. Building one takes two to four weeks. For a single deal.

Then there are broker opinions of value. Valuation models. Market analysis reports. Lease comparison spreadsheets. Every deal generates a small mountain of documents that someone has to create, format, check, and revise. In most brokerages, that someone is either the broker themselves or a back-office analyst who is already working on five other deals.

The numbers are striking. Brokers commonly work 60-hour weeks. By Closera’s estimate, roughly 35 of those hours are spent on tasks that could be automated. That is not a marginal efficiency gain. That is more than half the workweek.

The existing software in CRE is not solving this. CoStar and Yardi dominate the data and property management layers. Buildout handles some marketing material creation. Reonomy and Cherre focus on data analytics. But none of them function as an AI workforce that actually produces the deal materials from end to end. They provide data. They provide templates. The broker still has to do the assembly.

The proptech market has seen plenty of investment but most of it flows into residential real estate, which is sexier and more consumer-facing. Commercial real estate technology is underfunded relative to the size of the market. Global commercial real estate is a $40 trillion asset class. The technology serving it feels about a decade behind.

The Micro: Stanford CS to CRE Back Offices

Closera was founded by Anmol Tukrel and Chinmay Patel. Both are Stanford Computer Science graduates who met in their first CS class. They came through Y Combinator’s Summer 2025 batch and are based in San Francisco. Two-person team.

Anmol’s background is at the intersection of AI and product monetization. He worked at Google on Gemini, NotebookLM, and Flow. Working on AI products at Google scale gives you a specific understanding of how to build AI systems that are reliable enough for professional use, not just impressive in a demo. Chinmay comes from BCG, where he advised Fortune 500 companies on AI automation, and previously worked at Point72 and Roblox. The consulting background means he has seen how large organizations actually adopt technology, which is relevant when your customers are commercial brokerages that are not exactly early adopters.

The product automates the creation of deal materials that brokers currently spend weeks on. Sales decks that took four weeks now take minutes. Property valuation models that took hours happen in seconds. Data verification across thousands of listings runs simultaneously instead of consuming days of manual checking.

I like the framing of “AI employees” rather than “AI tools.” It signals that Closera is not building a feature to bolt onto existing CRE software. They are building autonomous workers that handle entire workflows. The difference matters for pricing and positioning. Tools get compared to other tools on price. Employees get compared to the cost of hiring people.

The competitive landscape in AI for CRE is still early. Buildout does marketing material creation but it is template-driven, not AI-native. Skyline AI was acquired by JLL in 2022 and focused on investment analytics, not deal material production. Enodo does multifamily valuation. Dottid handles deal management. None of them are trying to automate the actual creation of offering memorandums and broker opinions of value. Closera is going after the production layer that sits between data and deals.

The Verdict

I think Closera is going after a real pain point in an industry that is ready for this. Commercial brokers know their workflow is inefficient. They just have not had a product that addresses the specific bottleneck of deal material production.

The 35-hours-per-week automation claim is bold. I would want to see that validated with real brokers on real deals before taking it at face value. Even if the actual number is 15 hours, that is transformative. A broker who gets 15 hours back per week closes more deals. The ROI math writes itself.

The risk for a two-person team is the depth of domain knowledge required. Commercial real estate has terminology, conventions, regulatory requirements, and market-specific norms that vary by property type, geography, and brokerage. An offering memorandum for a retail strip center in Dallas looks different from one for a Class A office building in Manhattan. The AI needs to handle that variation without producing documents that a seasoned broker would immediately flag as wrong.

In 30 days I want to see three to five brokerages actively using the product on real deals. In 60 days I want to compare a Closera-generated offering memorandum against one produced by a traditional analyst and see if brokers can tell the difference. In 90 days the question is whether this is a single-player tool or whether entire brokerages are adopting it as standard workflow. The former is a nice business. The latter is a category-defining one.