Customer: Loan Vision
Website: https://www.loanvision.com/
Country: United States
Industry: Financial Software / Mortgage Banking Technology (ISV)
Products and services: Microsoft Copilot Studio, Microsoft Azure (Azure Functions, Azure Container Apps, Application Insights), Microsoft Dynamics 365 Business Central
Loan Vision is the financial platform of choice for independent mortgage banks. Built on Microsoft Dynamics 365 Business Central and purpose-built for the way mortgage lenders actually operate, it serves nearly 300 lenders across the United States and is backed by a team that knows the industry as well as anyone in it. That depth is exactly why Loan Vision saw, earlier than most, that AI was about to reshape how its customers worked. The company set out to put a trusted AI expert directly inside its product. To build it at the level the vision demanded, Loan Vision went looking for the most AI-capable partner it could find and naturally landed on Volt Technologies.
Volt is an AI-first Microsoft partner, a 10x Microsoft Inner Circle member (among the top 1% of partners worldwide), and the engagement played to exactly that strength. In a little over four months, Volt designed and shipped LV Luna, a production AI agent embedded natively inside the Loan Vision experience. LV Luna answers complex, mortgage-specific product questions in about a minute, draws on a freshly structured knowledge base spanning Loan Vision's entire product, and lives right inside Business Central instead of in a separate window. With it, Loan Vision became one of the first Business Central ISVs to ship a production AI agent inside its own product, and the early results have been strong enough that the two teams are already building the next chapter together.
That dual fluency showed from the very first conversations. "They came in with both an AI perspective and Business Central expertise versus just a typical implementation approach," Amanda said. "Positioning it to start as building a new product versus just implementing a system was a huge highlight for the decision-making process." Volt ran the engagement like the product build it was. The first weeks were spent learning Loan Vision's business in real depth and capturing the scenarios LV Luna would need to handle. From there the teams moved through iterative design and customer-led "gold scenario" show-and-tells, where Loan Vision posed the real-world questions its users actually ask and validated LV Luna's answers against the responses they knew were right. The approach landed immediately as a rigorous, AI-first way to build something the team could trust.
The first thing Volt did was unlock Loan Vision's expertise. Working hand in hand with Loan Vision's specialists, Volt built a purpose-made documentation agent using GitHub Copilot, defined a consistent structure for every one of Loan Vision's modules, and turned decades of product knowledge into a clean, structured library spanning the equivalent of roughly 10,000 pages. That library became the brain behind LV Luna. Volt then served it to the agent through a custom tool built as an Azure Function, holding the documentation, product logic, and Business Central source code together in one place, and engineered the retrieval so LV Luna reasons over far more relevant context per question than a conventional approach would surface. The result is answers that are not just fast but genuinely good. As Volt's Solution Lead, Gonzalo Rios, framed the priority, Loan Vision "doesn't just want something that's fast, they want something they can trust to deliver real value," and that is exactly what the architecture delivers.
Then Volt made LV Luna feel like a true part of the product. Where most partners would have stopped at a Copilot Studio agent and left users to chat with it in a separate tool, Volt's engineers built a polished chat experience directly inside Business Central, connected to Copilot Studio through the Microsoft Agent 365 SDK, so users get help without ever leaving their work, and made the same agent available in Teams. They designed two intelligence tiers, giving internal Loan Vision teams deeper, source-level answers and historical support context while external customers get clear product and configuration guidance. They built an adoption dashboard on Azure Application Insights and Container Apps so Loan Vision can see exactly how LV Luna is being used. And to put LV Luna in the hands of nearly 300 customers smoothly, Volt built a self-service deployment wizard that handles the technical setup securely in the background. The whole solution runs on Copilot Studio, which gives Loan Vision and its customers enterprise-grade Microsoft security and governance by default. It is end-to-end product thinking that very few partners in the Business Central ecosystem bring, and it reflects how Volt works. As Gonzalo put it, "we don't just develop something. We go beyond the development when we partner with you."
“I can't say enough amazing things about the Volt team. They are incredibly organized. I love that they are dedicated to our project, so they're not pulled in different directions. And I've never worked with a vendor or partner who makes themselves so available.”
Amanda Blake,
Product Owner, Loan Vision
There was a moment, partway through the build, when the Loan Vision team realized what they had. In a show-and-tell, one of the company's VPs started throwing harder and harder questions at LV Luna, trying to find its edges. "He immediately began to light up," Amanda remembered. "From that moment forward, you knew this was going to be an incredible resource for our team." The proof points kept coming. When a team member challenged LV Luna with a commissions question, asking what loan amount at a given rate would produce a specific payout, the agent ran the calculation backward, returned the answer, and then laid out the steps to set it up in the system. "Way above and beyond what we thought we would be able to use it for," Amanda said.
Inside Loan Vision, LV Luna changed how the whole company works. The support team became experts across every topic instead of leaning on a few specialists, which freed the implementation team to stay focused on their own projects. "Instead of one person being an expert on commissions, another on Continia, they're all now experts," Amanda explained. The impact spread well beyond support. Sales reps now answer client questions live on the phone, the marketing team uses LV Luna to learn the product and build guides, and new hires get up to speed with an always-available expert beside them. LV Luna can turn an unfamiliar topic into a complete guide with prerequisites and troubleshooting, walk a user through building an import from an uploaded file, read an error message and tell a development team exactly what to fix, and surface reporting from connected systems that leaders could not easily get to before. As Gonzalo described it, LV Luna does not just hand users a tool, it shows them what they can do with it, deepening how well people understand both Loan Vision and the Business Central platform underneath.
The early customer response told the same story. Across a two-week beta with early adopter mortgage banks, feedback was overwhelmingly positive. Customers could resolve questions in the moment instead of pausing their work and waiting for a reply, which matters most during the month-end, quarter-end, and year-end crunch. They taught themselves parts of the system they had never adopted, and more than a few said LV Luna's guidance went past what they expected. One beta user at a national mortgage lender noted that LV Luna "provided additional insight and things to check that were not covered by support." Another, needing to pull a full list of branches with no idea where to start, said the agent "asked me one clarification question, then gave me step-by-step instructions. Easy." The most telling reaction was the simplest. As Amanda relayed, one customer told the team their staff used it every day and "I hope you're not going to take this away. We will be upset if you take it away." LV Luna did exactly what Loan Vision set out to do. It turns customers into confident product experts and positions Loan Vision as a forward-thinking, AI-enabled leader in the mortgage banking industry.
“We're really, really pleased with this partnership so far. Because of our progress, and just a great track record we have in this partnership thus far, I think we're headed in the right direction.”
Paul Loftus,
Chief Executive Officer, Loan Vision
Loan Vision's platform is deep, and the people behind it know mortgage banking cold. That expertise is the company's biggest asset. It is also, by its nature, concentrated. Years of hard-won knowledge about commissions, loan schemas, month-end processes, and the hundred small decisions that make a mortgage bank run lived largely in the minds of a handful of specialists and across notes kept in many different places. As Loan Vision's Product Owner, Amanda Blake, put it, "everyone had their own ways of documenting their knowledge, so it wasn't all in one centralized location." For a specialized product built by domain experts, that is a natural stage of growth. It also meant the company's best knowledge was not yet working as hard as it could.
The strain showed up first in support. A three-person team handled roughly 4,200 cases a year, often leaning on the deepest experts to answer the hardest questions. Many of those experts also sat on the implementation team, so tough tickets pulled them off active projects, and response times depended on who was free. The same depth that made the product powerful made it a lot to learn. Amanda, who had even run Loan Vision at a previous company, estimated she knew only about 30% of the system when she joined. "There is so much to Loan Vision," she said, and getting all of that capability into the hands of every employee and every customer was the real opportunity.
So, Loan Vision made a deliberate, forward-looking bet. The mortgage market was moving toward AI, customers were starting to expect it, and very few Business Central ISVs were actually embedding it into their products. Loan Vision wanted to be one of the first, and it wanted to do it in a way its security-conscious customers could trust. Knowing mortgage banking inside out, the team had the industry depth to know exactly what the agent needed to do. What they wanted alongside that was a partner fluent in both worlds, one with deep Business Central knowledge, genuine AI expertise, and the engineering firepower to match the ambition of the idea. They found it in Volt.
The success of LV Luna gave both teams the confidence, and the foundation to go further. With a trusted knowledge base in place and customers already relying on the agent every day, Volt and Loan Vision are now building LV Luna 2, which extends everything LV Luna does today into action. The same agent that explains how to post a loan or build a commission profile will be able to do it, working directly and securely inside each customer's own Business Central tenant. Volt deliberately built the product in this order, earning customer trust and locking in security first, so that the move to action stands on a proven base rather than starting from scratch.
The vision is for an agent that is even more proactive, not just reactive. As Amanda described it, LV Luna 2 could review incoming loans and flag the ones that are out of balance before anyone asks or takes on the deprioritized tasks that pile up in busy accounting teams. For Loan Vision's customers, that means doing more with the same team and confidently adopting powerful modules, like commissions, that some hold back on today because the setup is involved. It is the natural next step for a product that has already proven its value, and it reflects the way Volt partners: not handing over a finished build and walking away but staying alongside the customer as the product and the technology keep evolving. Loan Vision's leadership is already looking ahead to it. "We're really excited about Luna V2," said Chief Operating Officer Ben Saunders. "I think there's a lot of value we're going to be unlocking there."