LoanForge
An AI mortgage loan-origination platform that extracts borrower data from documents, auto-clears underwriting conditions, and drives every loan to clear-to-close.
// The Loan Pipeline board — cycle-time, auto-clear and pull-through KPIs, a cycle-time trend falling to 14 days, and the active-loans table sorted by conditions outstanding.
Figures from the engagement, shared under NDA.
We publish them unchanged and can’t break them down any further here. The reference build you can open runs entirely on sample data.
If you read one part of this page, read these 4 lines.
- Turns a stack of borrower documents into extracted, source-cited loan data
- Auto-clears 68% of underwriting conditions without a human touch
- Drafts the exact borrower request for whatever is still outstanding
- Every extracted field carries a confidence score and cites its source document
What it does — and which part of it we built.
LoanForge turns a folder of borrower documents into a clear-to-close decision.
It reads income, employment, assets and liabilities from W-2s, paystubs, bank statements and appraisals — each field tagged to its source page with a confidence score — reconciles them against the 1003 and agency guidelines, and auto-clears the underwriting conditions that pass every rule. BuildspaceLabs designed and built the MVP front end — a Loan Pipeline board for the whole book and a single loan file where extraction, conditions and ratios come together — turning a document-heavy origination workflow into one surface a loan officer can drive to close in days, not weeks.
Every screen here is from the reference build, not the client’s system.
The captions say what you are looking at. Every name and record on them is sample data.
What the engagement left the client with.
- Delivered a production-quality MVP of the Loan Pipeline board and loan file view in a 10-week engagement
- Consolidated document extraction, condition-clearing and qualification into a single origination workflow
- Gave the origination desk a real-time picture of cycle time, auto-clear rate and pull-through across 312 loans
- Established a reusable design system (loan-type pills, stage funnel, confidence rings, condition checklists) for the product's next surfaces
The stack was picked per constraint, not per house habit.
7 technologies across 3 layers — the shape of the stack follows the problem.
Every feature in the build (6)
- Loan Pipeline board with cycle-time, auto-clear and pull-through KPIs above a colour-coded active-loans table
- Document extraction of income, employment and assets — each field with a confidence score and its source document
- Auto-clearing of underwriting conditions that pass every rule, with exceptions routed to a human
- AI-drafted condition requests for outstanding items, with a confidence score and ready to send
- Live DTI, LTV and PITI recomputed from extracted income and liabilities, checked against guidelines
- Loan file detail with an extracted-fields panel, conditions checklist, ratios and next action
Don’t take our word for it — it is running in the frame below.
This is the same page the provenance block links to: our reference build for LoanForge, with its own look, copy and motion. Scroll inside it, or open it full-screen.
// next
Building something in this space?
A 30-minute call is enough for us to understand your situation and tell you whether we're the right team. Usable version in the first two weeks if we are.