API-first workflow core for mortgages
Mike Yu, CEO of Vesta, on building a new system of record for the mortgage industry
This is really a claim about control, not just old software. In mortgage origination, the core loan origination system sits in the middle of every step, from application intake to underwriting to closing, so when that core is closed, every new tool has to ask permission to connect. That turns basic integrations into long partner projects, paid services work, and vendor dependence, which is why a modern API first core matters so much.
-
The legacy setup was built around a loan file that humans pass around and recheck repeatedly. Vesta's argument is that newer software should break that work into smaller tasks and rules, so the system can route work automatically instead of making each processor or underwriter inspect the whole file again.
-
The walled garden dynamic shows up in practice through controlled partner access and legacy integration frameworks. ICE's Encompass requires partner setup, lender specific API keys, and works within platform limits, while also still pushing customers through a multiyear sunset of older SDK and legacy integration paths.
-
The market has been filling in around these closed cores with point solutions, not replacements. Snapdocs handles closings, Polly handles pricing, Blend and SimpleNexus handle borrower facing flows, but those products still need a system in the middle that can pass data cleanly across vendors instead of forcing lenders into one bundled stack.
The direction of travel is toward a modular mortgage stack, where the core becomes a workflow engine and data hub rather than the vendor that owns every adjacent product. As lenders keep replacing manual steps with software, the winning core will be the one that makes outside tools easy to plug in, not expensive to ask for access.