SBA Loans HQ is a US-based consulting firm that helps businesses secure SBA financing. When the project kicked off, the entire operation ran on email threads, spreadsheets, and a collection of disconnected SaaS tools. Brokers were chasing documents manually. Lenders had no visibility into deals. Nothing talked to anything else.

They needed one platform to run the whole business. A team of four built it from the ground up. Six months later there was a production system handling real deals, real money, and real users, with 50+ lenders onboarded and 200+ deals actively moving through the pipeline.

Three things stood out after understanding how the business actually worked.

Brokers were losing deals in their inbox. Document requests went out over email, follow-ups were manual, and nothing was tracked in one place. High-value deals were slipping through because of basic coordination failures.

There was no pipeline visibility. Leadership had no way to see where deals were stuck, which partners were performing, or where the bottlenecks were. Every status update required someone to ask someone else.

Third-party tools were duct-taped together. Zapier, Copper, and a handful of other tools were handling critical business logic. It was fragile, expensive, and impossible to customize.impossible to customize.

A multi-role platform supporting four user types: partners, borrowers, lenders, and external requestees. Each role gets a tailored experience with the right access and the right tools.

Role-based team access - Coordinators, managers, and admins each see only what's theirs, coordinators get their assigned deals, managers see their team's pipeline, admins see everything. Per-user activation, last-login tracking, and deal assignment all keyed to role. The audit trail is built into the access model itself, not bolted on after.

Branded partner workspaces - Every partner that joins gets a white-labeled workspace, their own apply form, custom loan stages, qualification rules, e-signature templates, and a partner memorandum. The apply form ships in three embed formats (direct URL, iframe, or JS widget) so partners drop it straight into whatever marketing site they already run. Onboarding a new partner used to be a week of manual setup. Now it's under a day.

Lead intelligence - Inbound leads used to pile up in brokers' inboxes, sorted manually. Now every lead is scored the moment it lands, AI Agent reads credit signals, loan purpose, business type, and behavioral cues, then routes high-intent prospects straight to a broker. Cold leads go through automated nurturing without ever touching a human. Brokers wake up to a sorted queue instead of a triage list.

Deal pipeline - A Kanban and table view that tracks every loan from first contact to approval. Each stage automatically triggers the next set of document requirements, so brokers never have to remember what to ask for next. Doc-completion rings on every card, last-login signals, and stuck-deal flags surface what needs attention before a deal goes cold.

One of the bigger time sinks in SBA lending is the back-and-forth over what documents are needed and when. Every broker had their own way of asking, which meant inconsistent requests, missing items, and deals stalling because borrowers did not know what to send next.

The platform solves this with a bundle system. Partners build reusable templates that group together the exact documents and questions needed at each stage. When a deal reaches a new stage, the right bundle goes to the borrower automatically. No manual follow-up, no copy-pasting the same email for the hundredth time. A process that used to take days of back-and-forth now completes in hours, with full visibility on both sides.

Borrowers get a clean, guided experience on their end. They see exactly what's being asked, can upload files directly, and answer questions inside the platform. Partners get notified when items come in and can see at a glance what's still outstanding per deal.

The result is a process that used to take days of back-and-forth now completing in hours, with full visibility on both sides.

Document vault - Secure file management backed by AWS S3. Documents are automatically bundled and requested based on where a deal sits in the process. DocuSeal integration lets borrowers sign directly inside the platform without going anywhere else.

AI document processing - AWS Textract pulls structured data out of every financial PDF that comes in — tax returns, bank statements, P&Ls, the whole stack. What used to take 30 minutes of manual entry per document now happens in seconds. Brokers stop typing and start reviewing.

In-platform messaging - Every deal has its own chat thread. Brokers, borrowers, and lenders communicate inside the platform, attached to the deal they are discussing. Email is no longer part of the workflow.

Live analytics dashboard - A live view of the entire operation for super admins, covering lead distribution, partner performance, deal velocity, and where things are getting stuck. Every metric refreshes the moment a card moves on the pipeline. Leadership stops asking for status reports because the report is always on screen.

Custom automation engine - All third-party connectors were replaced with a backend automation layer built in-house. It handles stage-based triggers, document bundle creation, and outbound communication via Postmark. More reliable, cheaper, and fully in one place.

$33M+ in loans approved through the platform

50+ lenders onboarded

100+ active deals in the pipeline

25% increase in annual revenue

60% reduction in operational overhead

30% faster loan processing times

6 months from kickoff to production

Frontend: Next.js, Tailwind CSS, ApexCharts

Backend: Django REST Framework, PostgreSQL, AWS RDS

AI: Gemini AI, AWS Textract

Infrastructure: AWS EC2, S3, Lambda

Integrations: Postmark, DocuSeal, Mixpanel, GA4