Dark Matter Technologies takes on the waiting.

At Lake Michigan Credit Union, twelve people once monitored and ordered verifications and documents. After the lender adopted Empower, that work required one person. Experienced employees moved into other roles. In the company’s case study, the striking detail is how much skilled attention had been absorbed by keeping paperwork moving.

This is the territory of Dark Matter Technologies: the space between receiving a mortgage application and getting a loan ready to close. Its customers are banks, credit unions and independent mortgage lenders. Its proposition is that a great deal of mortgage work can move forward without another person remembering to push it.

A startup created by a merger

Dark Matter’s origin is unusually revealing. In 2023, regulators challenged Intercontinental Exchange’s acquisition of Black Knight. ICE already owned Encompass, a major loan origination system. The Federal Trade Commission argued that combining it with Black Knight’s Empower would reduce competition in the software lenders use to make mortgages.

The settlement required divestitures. Constellation Software’s Perseus group acquired the Empower business, and Dark Matter launched that September under Rich Gagliano. The stated Empower transaction price was $40 million, subject to adjustments. That bought an existing business. It was separate from the $700 million Optimal Blue transaction.

So a company born in 2023 could arrive with established software and longstanding customer relationships. Its competitive position had a practical foundation: lenders still needed a separate alternative to Encompass. Keeping that alternative useful meant carrying forward familiar products while changing what happened inside them.

Follow the file, not the feature list

Empower is the operating center. It holds the loan record and coordinates tasks from application through closing. Its exception-based processing uses rules and changing loan information to determine what happens next. A file can generate a task for the right person when something needs attention, rather than depend on a sequence of manual handoffs.

AMOCO Federal Credit Union chose Empower alongside borrower and loan officer portals, plus connectivity to its servicing system. Its published case study describes automatic notifications and document classification. The lender reported that closing, previously taking 35 to 40 days, averaged 27 days. Those are results from a particular implementation, not a forecast for every lender.

LMCU’s initial trigger was different: its former system struggled to keep up with regulatory changes. It wanted compliance support and more flexibility through mortgage cycles. That is a useful corrective to the usual software shopping list. A faster screen matters less when the underlying process keeps producing work that has to be corrected.

The business is software as a service, accompanied by configuration, training, implementation and continuing support. Buyers are selecting an operating process as well as a product. Dark Matter describes a phased rollout with professional services and account management. The sensible procurement exercise is to price that whole relationship against the work it can actually remove.

The pay stub gets an answer

Aiva supplies the document intelligence. It classifies files, extracts information and supports income, asset and quality reviews. Embedded in Empower, it can feed information into the same workflows that route the next task. The attraction is less copying between tools and fewer occasions for a file to sit untouched.

In August 2026, Dark Matter extended validation to the moment a borrower uploads a document. Consider a pay stub submitted against a condition requiring a particular period of income. The system checks the document against that condition and returns an explanation. A borrower can discover a problem while still looking at the application.

Ask Aiva, introduced in April 2026, tackles another delay: finding an answer inside the origination environment. It retrieves relevant information and produces a conversational response with access to supporting sources. That traceability matters when an employee must understand the basis for an answer before acting on it.

“The best idea has to win”Vikas Rao / September 2026 CEO letter

The portfolio reaches beyond origination. Elevate handles servicing, including payments, reporting and borrower self-service. Exchange connects lenders to outside service providers. In October 2026, Dark Matter announced a Mezzo integration for routing verification requests across vendors through a single gateway. The recurring design problem is getting information to its next destination without starting over.

For an institution with fewer technology staff, the company also added NOVA LOS to its lineup in 2024, positioning it around simpler administration. For lenders building their own connections, the Developer Platform publishes API documentation and downloadable specifications. This gives the portfolio two routes into the market: institutions that want a configured operating system, and teams that want room to build around one.

Who holds the dial?

Dark Matter’s expertise combines mortgage operations, product development and the integrations around a lender’s core system. Chief product officer Stephanie Durflinger previously worked at ICE, Sagent and Ellie Mae. CEO Vikas Rao, appointed in April 2026, also brought experience with Encompass and developer ecosystems. These are people familiar with the machinery their customers are being asked to replace.

There is also an organizational cost to changing direction. The CEO transition announcement included a reduction in force. Later, Rao described a culture of clarity, urgency and accountability. Those ambitions belong alongside the restructuring when judging the company’s evolution.

For lenders, the crucial control is how automation gets supervised. Document validation can advance qualifying files automatically or route them to a loan officer. Configuration decides the boundary. Weak rules, poor inputs or an overloaded exception queue can simply move the bottleneck. Software still needs people who know what a satisfactory result looks like.

The lesson another business can copy is to follow one piece of work. Count its stops. Ask which checks could happen when the information arrives. Give the exceptions an owner. Dark Matter’s case rests on what happens after that exercise: fewer people chasing the next step, and more attention available for the borrower who needs it.