Explainable credit scoring
Analyze financial, behavioral, market, and macro signals with score drivers, benchmarks, and traceable decision reasoning.
Credit-analysis software for lenders, with configurable origination, reconciliation, and ledger-disclosure workflows.
Truffles AI provides credit-intelligence software for lenders and financial teams. Its scoring engine combines financial statements, alternative data, and market signals with factor-level explanations and decision trails.
Its enterprise systems cover document processing, loan origination, servicing reconciliation, and general-ledger disclosure preparation. Workflows are configured around the client's data and operations. The Orchid case study shows a lender-specific deployment with a shared loan record across browser, mobile, and API interfaces.
Analyze financial, behavioral, market, and macro signals with score drivers, benchmarks, and traceable decision reasoning.
Use financial statements, PDFs, spreadsheets, and system exports in a pipeline that normalizes data, validates it, and tracks sources.
The Orchid deployment classifies scanned files, resolves borrower and co-signer records, extracts fields, and advances files through controlled approval and servicing stages.
In the loan-system case study, expected payments are compared with remittance files at borrower and period level, with ambiguous documents or exceptions sent for human review.
Map ledger data from an ERP replica into disclosure schedules, preserving source-to-output lineage and reviewer approval for material outputs.
Review borrower financials and model explanations before making or presenting a credit decision.
Use a demonstration to assess how document intake, borrower records, approvals, and reconciliation would fit the lender's own operations.
Map source ledger records to workpapers and disclosure schedules while retaining human review of material outputs.
Use the public loan-pricing calculator to examine break-even rates, target ROE, and default-rate stress under its simplified assumptions.
Truffles offers demonstrations tailored to lender data, loan files, or ERP and disclosure requirements, plus a product sign-in portal. The public calculator is separate from access to the enterprise systems.
No. Orchid is a named client deployment. Its native iOS app is for authorized employees of the client organization, and its browser and API surfaces share that organization's loan records.
It is an estimate using a single-period, pre-tax model. It holds exposure flat instead of amortizing the loan, treats capital as a fixed share, and spreads the origination fee across the term. Funding, default, recovery, and capital assumptions affect the results; it is not a definitive lending quote.
No. Ambiguous loan documents and reconciliation exceptions are routed to people, and GL workflows include review of material outputs. The GL page lists SOC 2 Type II as in progress, rather than a completed certification. Model and filing outputs still require the organization's judgment.
GL supports deployment in the client's private cloud. The privacy policy allows processing by trusted service providers, and client mobile apps send loan documents to the client's systems. Data and model-retention arrangements depend on the chosen deployment and agreement.
Os dados de tráfego são apenas para referência.