Integrations

Vyaso runs on the data your stack already produces.

Whether you parse with Ocrolus, Heron Data, Validis, or upload raw PDFs and CSVs, Vyaso reads the output and re-underwrites it. We are not the parser. We are the layer that re-underwrites what the parser produced.

The two layers, clearly.

The parser layer

What is each transaction?

Tools like Ocrolus, Heron Data, and Validis read the raw bank statement, extract transactions, normalize dates and amounts, categorize merchants, and produce clean structured data.

Transaction extraction · Description normalization · Counterparty resolution · Category labels

The underwriting layer

Do these transactions add up to revenue you can fund against?

Vyaso reads the parser's output and re-underwrites it, surfacing loan stacking, account kiting, circular flows, daisy-chain inflation, and forgery indicators. Then produces an adjusted-revenue figure and a risk score.

Loan stacking detection · Account kiting · Daisy-chain inflation · Adjusted revenue · Executive summary

Run them together. They answer different questions.

Side by side

The same statement, two outputs.

What a parser produces, and what Vyaso produces from it. On the same bank statement.

Parser output (typical)

04/03 ACH CREDIT ACME CORP +$48,000 Income
04/12 TRANSFER FROM SAVINGS +$30,000 Transfer
04/19 DEPOSIT QUICKFUND CAPITAL +$42,000 Income
04/24 DEPOSIT VELOCITY LENDING +$38,000 Income
04/27 CHECK 1284 PAYROLL −$12,000 Payroll
04/30 ACH DEBIT NET-30 SUPPLIER −$8,000 Expense

Clean, structured, normalized. Six rows of data. No verdict on what they mean.

Vyaso output (same statement)

Risk tier:High
Adjusted revenue:$48,000 of $158,000 gross

Top flags (3):

  • · Loan stacking (2 lender deposits, 5 days apart)
  • · Account kiting (transfer from savings)
  • · Lender deposit not customer revenue (Velocity)
Network graph:4 counterparties, 1 cycle detected

Same six rows. A verdict the credit committee can defend.

Run them together. The parser answers what each transaction is. Vyaso answers whether the merchant should be funded.

Run Vyaso on your existing pipeline.

Pilot-only
Ocrolus

Ocrolus → Vyaso

Pull normalized transactions from Ocrolus directly into Vyaso. Run the nine layers and the synthesis step on their structured output, with no separate upload.

Talk to us about Ocrolus
Pilot-only
Heron Data

Heron Data → Vyaso

Read Heron Data's categorized transaction stream into Vyaso. Use their merchant categorization as input to Vyaso's revenue confidence and counterparty analysis.

Talk to us about Heron
Pilot-only
Validis

Validis → Vyaso

For funders pulling bank-account data through Validis, Vyaso reads the financial-data API and re-underwrites the resulting transactions.

Talk to us about Validis
✓ verified
Raw upload

PDF / CSV upload

The default, no upstream parser required. Drop in the bank statement PDFs (or a CSV from your data warehouse) and Vyaso parses, categorizes, and re-underwrites in one pass.

Request a pilot

Plug into your stack.

REST endpoints

✓ verified

Upload, retrieve results, export transactions, fetch the network graph data, all via standard REST endpoints.

CSV / JSON export

✓ verified

Every analysis is exportable in either format. Drop into your data warehouse or attach to the submission packet.

Webhook on completion

Roadmap

Subscribe to analysis-complete events for asynchronous pipelines.

SSO and per-user accounts

Roadmap

Single sign-on, role-based permissions, audit logs.

White-label dashboard

Roadmap

ISO-branded report exports and dashboard themes.

Frequently asked.

Run Vyaso on your portfolio.

Free 30-day pilot. Bring 50–100 files. We'll show you what the model would have flagged, what would have been approved, and how the adjusted revenue compares.

Request a pilot

No commitment. No setup fee.