Regulatory Compliance & Document-Intelligence Platform

Prisma — Regulatory Request Processing

• Ingests official information requests ("oficios") from Mexican banking regulators regarding disputed customer transactions.
• Extracts and validates document data through an OCR and document-intelligence pipeline.
• Reconciles extracted data against internal records and generates regulator-compliant response documents.
• Validated end-to-end against a purpose-built simulator of the regulator's submission portal.

IndViSi — Statement Verification

• Automates the quality check banks are required to perform on customer account statements before issuance — arithmetic accuracy, customer data, formatting and required legal disclosures.
• Applies a configurable, bank-specific rules checklist to each statement and produces a findings report.
• Flags exceptions for human review rather than making a final accept/reject decision on its own, keeping a qualified analyst in the loop for regulatory judgment calls.
• Built around a statement-verification engine — Verificación de Estado de Cuenta in Spanish.

Architecture

Both products share a common hexagonal-architecture .NET foundation and document-processing pipeline, with IndViSi built as an additional module on top of the platform originally developed for regulatory-request processing. Ingestion, processing and validation are separated into independent stages, using a functional error-handling model (Result<T>) in place of exceptions for business logic, with full observability instrumentation throughout.

Technology

• .NET 10 / C#, hexagonal / clean architecture
• Blazor (MudBlazor) web interface
• OCR and document-AI processing (Python interop)
• Event-driven processing (Rx.NET)
• xUnit v3, Serilog, OpenTelemetry