Dark Matter acquires House Numbers to automate mortgage lending

The acquisition details

Dark Matter Technologies recently announced the acquisition of the artificial intelligence platform and intellectual property belonging to House Numbers. House Numbers was a venture capital-backed company that specialized in creating AI agents for the mortgage industry. The deal brings both the technology and the core team from House Numbers into Dark Matter's existing home lending business.

The leadership transition is a key part of the deal, with House Numbers co-founder and CEO Jeff Levinsohn becoming the chief AI officer at Dark Matter. Additionally, co-founders Arthur Granado and Benjamin Hoffman will join the company as vice president of engineering and vice president of AI adoption and enablement, respectively. This move ensures that the expertise behind the original platform will directly guide the integration of these tools.

While the financial terms of the transaction were not disclosed, the move was intended to move beyond a simple partnership. Dark Matter CEO Vikas Rao explained that he proposed the acquisition to ensure the technology would be deeply integrated into their products rather than offered as a standalone service. This strategy allows the technology to operate at a larger scale than it could as an independent entity.

How agentic AI works

The technology being acquired is built on 'agentic' AI, which refers to autonomous agents capable of completing specific tasks rather than just generating text. These agents are designed to monitor loan files for changes and work through various loan conditions automatically. They can also communicate directly with borrowers and lenders using email, text, and voice messages.

Currently, the platform has 19 agentic AI use cases in production, with more applications still in development. These tools can review incoming documents to see if loan conditions have been met, check closing packages for missing signatures, and flag potential issues for human review. They are also designed to help borrowers understand what requirements they still need to fulfill and provide feedback on incorrect documents.

To ensure reliability and safety, the system is designed to run within Amazon Web Services infrastructure with specific controls for security and regulatory needs. Lenders have the ability to determine exactly when an agent can act independently versus when it should only make recommendations. Furthermore, the system allows lenders to train these agents to follow their specific internal policies and procedures.

Impact on mortgage costs

One of the primary goals of this acquisition is to lower the significant costs associated with originating a mortgage. Vikas Rao noted that the total cost of producing a loan is currently around $12,000, with approximately $6,000 of that amount attributed to operational expenses. By using AI to handle repetitive tasks, Dark Matter hopes to shave a significant amount off that $6,000 figure.

The company believes that even a $1,000 reduction in operational costs would be a major shift for the industry. By bringing the full suite of House Numbers' capabilities to market, they expect to exceed those types of savings. The technology is intended to perform tasks instantaneously, which allows for a faster flow of information between the lender and the borrower.

Despite the automation, the company emphasizes that the goal is not to replace human workers entirely. Rao stated, "We’re not talking about eliminating humans," and clarified that the AI agents perform tasks quickly while a human still reviews the final output. This hybrid approach aims to provide a safety net while increasing the speed at which loans move through the origination process.

Product integration and roadmap

Dark Matter plans to weave this new technology into its existing loan origination system and its built-in point-of-sale platform. The company already offers other products, such as Aiva Intelligence and the Elevate loan servicing platform, which will also see updates. The House Numbers team will work across these various products to enhance the tools already in production.

Customers can expect to see new capabilities appearing within weeks of the announcement. Following that initial rollout, the company has planned additional releases over the coming months to further expand the AI's reach. The goal is to provide a more seamless experience where AI assistants help loan officers identify better options and track key dates without navigating multiple screens.

Looking ahead, Dark Matter intends to continue pursuing acquisitions to grow its portfolio. The company's parent organization, Constellation Software, encourages its portfolio companies to seek out complementary acquisitions. Rao indicated that the company is looking at several other potential companies in the market and hopes to pursue more deals in 2027.