MCA underwriters spend hours per deal manually reviewing bank statements and researching merchants. Draco AI automates the analysis so you can focus on the decision.
Every MCA deal requires the same manual grind. We built tools to eliminate it.
Draco Alpha
Draco Alpha turns your submissions into complete underwriting analysis, so your underwriters start at the decision instead of the data.
‹ All submissions
Pizza Restaurant
(212) 555-0148
123 Main St, New York, NY 10013
Family-owned pizzeria serving dine-in, takeout and delivery since 2014.
Est. monthly revenue
$32,015
Est. monthly expenses
$30,942
Net
$1,073
Avg. daily balance
$4,860
| Lender | Status | Payment | Frequency | Last payment | Funding amt |
|---|---|---|---|---|---|
| Active · 2 | |||||
| Vader Capital | Current | $331.00 | Daily | 01/30/2026 | $25,000 |
| Kestrel Funding | Current | $1,250.00 | Weekly | 01/27/2026 | $40,000 |
| Default · 1 | |||||
| Orbit Advance | Stopped By Merchant | $289.00 | Daily | 11/14/2025 | $15,000 |
| Paid off · 1 | |||||
| Summit Merchant | Paid In Full | $410.00 | Daily | 08/22/2025 | $20,000 |
Submission documents, scanned or digital, come out as structured data.
Each bank account is pieced back together across all its statements, even when they come from different sources.
Real revenue is separated from transfers, MCA funding and returned items.
Every existing advance is found, with who's being paid, how much, how often, and whether each one is current, in default or paid off.
The business is looked up on the web, and what turns up is checked against the application.
Automatically detect bounced and stopped payments on lenders, undisclosed transfer accounts, missing statements, and other risk flags.
Every deal moves through the same four steps, in Draco Alpha or through the API.
01
Upload the submission as it came in. Each document is automatically classified, inspected, and de-duplicated.
02
Quickly filter out bad deals with a fast, low-cost read on the most recent bank statement. If it's not your deal, don't underwrite it.
03
When the deal is worth pursuing, run full underwriting. Every document is extracted, every account rebuilt into a ledger and the business researched.
04
Read the report, trace any figure back to its account and statement, and model the offer.
Draco API
Bring submission underwriting and document extraction into your CRM, LOS or internal tools.
# Create a submission from uploaded PDFs
POST /mca/submissions
{ "file_ids": ["file-2a7f…", "file-91c0…", "file-c44e…"] }
# Run full underwriting
POST /mca/submissions/{id}/underwriting/jobs
# Read the results
GET /mca/submissions/{id}
{
"underwriting": {
"status": "completed",
"summary": {
"estimated_revenue": "32015.00",
"average_daily_balance": "4860.00",
"mca_positions": [ … ]
}
}
}Your financial documents are sensitive. We treat them that way.
AES-256 encryption for all stored documents
TLS 1.3 for all data transmission
Hosted on Google Cloud Platform
Your documents are never used to train models
The MCA industry processes millions of funding applications annually, yet most underwriting is still done manually. We chose to start here because we've done this work ourselves—we know exactly where the bottlenecks are and what underwriters actually need.
MCA underwriting is complex enough to matter, standardized enough to automate, and high-volume enough to deliver real ROI. It's the proving ground for underwriting AI that will eventually serve all of small business lending.
Domain expertise meets AI engineering. We've done the underwriting work ourselves.
We work with statements from virtually any US bank or credit union. We handle PDFs, scanned documents, and various statement formats.
We maintain a catalog of known MCA lenders and identify their payment patterns automatically. For lenders not in our catalog, we flag other recurring debt payments for your review. Our system continuously learns and expands its recognition as we encounter new funders.
Yes. The Draco API lets you add submission underwriting or document extraction to your CRM, LOS, or internal tools.
All documents are encrypted at rest (AES-256) and in transit (TLS 1.3). We host on Google Cloud Platform and never share your data or use it to train our models.
We charge based on what you process through our platform. Each product has its own unit and rate, so you only pay for what you use. Let's talk about what makes sense for your volume.
Our team works with you directly to get set up. We move fast and can typically have you processing statements within days.
Request a demo and see how Draco AI fits your underwriting workflow.
Request a Demo
