Antares Labs Raises $7.25M to Build AI for Real Estate

A Chicago-based startup just launched publicly with a specific bet: that real estate companies don't need another generic dashboard, they need AI built directly around their own data and workflows. Antares Labs raised $7.25 million in seed funding, with participation from Fifth Wall, Base10 Partners, Bloomberg Beta, and Sandwith Ventures, according to Commercial Observer's reporting on the launch. Fifth Wall, the largest investment firm focused on technology for the built environment, led the round.

Antares Labs is led by CEO Noaman Ahmad, who previously served as chief financial officer at proptech company Doma and technology consultancy Keystone. Ahmad described the gap he identified driving the company's founding directly: "Most of these companies have now realized that they need to do something using the power of AI, because if they don't, they're going to lose the competitive edge over the next year or two years."

The underlying problem Antares Labs targets is genuinely widespread across the industry. Large real estate owners and operators frequently possess decades of information about acquisitions, leasing, tenant behavior, property operations, and market conditions, but much of that knowledge remains scattered across spreadsheets, databases, and individual employees' heads rather than being usable by any system, according to Pulse2's reporting on the company's approach.

Rather than shipping a general-purpose software product customers must configure themselves, Antares Labs uses forward-deployed teams, technical and industry specialists who embed directly inside a customer's organization to learn how the business actually operates before building tailored AI capabilities around specific, high-value workflows. Real use cases already in production include shrinking the time between tenant departures and new arrivals for a shopping center-focused REIT, and generating current and future inventory insights for an industrial landlord, according to Mann Report's coverage of the launch.

A genuinely notable technical detail is Antares Labs' commitment to staying model-agnostic. The platform integrates with every major foundation model, including AWS Bedrock, OpenAI, Gemini, and Anthropic, dynamically routing requests to balance cost, performance, and quality rather than locking clients into a single AI provider, according to FinSMEs' reporting on the round. That flexibility matters given how quickly pricing and capability shift between competing AI labs, a dynamic worth understanding alongside our coverage of Kimi K3's recent open-weight release and the growing pressure it's putting on closed model pricing across every industry, including real estate.

Antares Labs is already working with Quarterra, an early customer demonstrating genuine enterprise adoption rather than pure pilot-stage experimentation, according to FinSMEs' coverage of the launch.

The forward-deployed engineering model Antares Labs uses is worth watching as a genuine differentiator in how AI gets implemented at scale. Businesses across many industries, not just real estate, have similarly fragmented institutional knowledge sitting in spreadsheets and individual employees' heads rather than usable systems. Antares Labs' approach, building AI around a company's actual existing workflows rather than asking the company to adapt to generic software, is a model worth evaluating for any business sitting on decades of unstructured operational data.

Antares Labs raised $7.25 million in seed funding to build custom AI systems for large real estate owners, developers, and asset managers, led by proptech investor Fifth Wall. The company uses forward-deployed teams that embed directly with clients to convert fragmented institutional knowledge into working AI systems built around specific business workflows. The platform is model-agnostic, integrating with AWS Bedrock, OpenAI, Gemini, and Anthropic to balance cost and performance rather than locking clients into a single AI provider.