Mortgage LOS Software Development

The client is a full-service mortgage lender based in the US, helping home buyers, owners, and investors purchase and finance homes. Eastoner has partnered with their internal team to modernize and extend their internal CRM and loan origination system (LOS).

The client's internal CRM ran on a legacy PHP backend alongside an early, unfinished .NET implementation. Technical debt was mounting, and the system couldn't support the third-party integrations the business needed for borrower verification and loan processing. The internal team lacked the capacity to handle a migration of this scope alongside ongoing feature work, so the client brought in Eastoner to migrate the backend and build it out into a full loan origination system, covering everything from lead intake to loan issuance or decline.

Legacy backend migration

Eastoner engineers migrated the backend from PHP to .NET Core, replacing the earlier partial .NET implementation with a properly architected system built on Web API and Asp.net, backed by SQL Server and hosted on Azure (Azure SQL, Azure Web App, Azure Key Vault for secrets management, Hangfire for background jobs). The migration cleared the technical debt blocking new features and gave the team a foundation to build the integrations the business needed.

Call logging and notification system

We built a call logging system integrated with the client's PBX that identifies both the loan officer and the borrower on a call when the borrower's information is already in the system. Calls are tracked automatically: the loan officer gets an in-app notification identifying who's calling, along with a reminder to start note-taking.

A notification system integrated with OneSignal handles missed-call alerts. Reminders can go out by SMS, push notification, or email, to loan officers and to customers, cutting the manual follow-up loan officers previously had to track themselves.

Automated borrower verification

Borrower verification through Fannie Mae and Freddie Mac was previously done manually. We automated it through Node Builder, which generates the required XML documents and sends verification requests for income, employment, and assets automatically. What used to be a manual, per-borrower process now runs without loan officer intervention for each request.

CRM, email, and calendar integrations

We built a HubSpot integration and a Microsoft Graph API integration for Outlook, including subscription and webhook management, so calendar and contact data sync without manual re-entry. We also developed a mass email feature the client's CEO described as exceeding expectations. The platform now integrates with LodeStar and Optimal Blue as well, supporting loan pricing and settlement data alongside the GSE verification flows.

Performance work

Ongoing performance tuning included a production test where task load time dropped from 1.3 minutes to 18 seconds. The team also built out unit test coverage and used Postman for API testing to keep the growing integration surface stable as new features shipped.

The legacy PHP/early-.NET CRM is now a modern .NET Core platform functioning as a full CRM and LOS system, handling the loan lifecycle from lead to issuance or decline for the client's loan officers.

Automated verification with Fannie Mae and Freddie Mac replaced a fully manual process. Call logging, notifications, and CRM/email integrations (HubSpot, Microsoft Graph, LodeStar, Optimal Blue) consolidated work that previously required loan officers to track manually across separate tools. One production test showed task load time improving from 1.3 minutes to 18 seconds.

The client continues to request new features, and the cooperation remains ongoing with the same 9-person team.