Legal & Financial · Miami, Florida

Sixty years of history, searchable in seconds.

Banco Núñez engaged Veritas Data to build a secure, AI-powered document intelligence platform for their Helms-Burton litigation strategy — transforming decades of historical banking records, correspondence, and financial documents into an instantly queryable knowledge base for legal counsel.

Engagement Summary
IndustryLegal & Financial
LocationMiami, Florida
ServiceDocument Intelligence Platform
OCR EngineAWS Textract
LanguagesEnglish & Spanish (EN/ES)
Archive Depth60+ years of records

Cuba's second-largest bank. A six-decade evidentiary challenge.

Context
Banco Núñez was Cuba's second-largest private bank before the Cuban government's nationalization of private industry in 1960. Under the Helms-Burton Act (Title III), U.S. nationals and Cuban-American claimants may pursue legal action against entities that traffic in confiscated property. Building a viable litigation strategy requires assembling and presenting decades of historical documentation to establish ownership, valuation, and chain of title.

The legal team faced a fundamental challenge: the evidentiary record for a Helms-Burton claim spans decades of banking records, correspondence, financial statements, board minutes, and government documents — many of them in Spanish, many of them scanned from physical originals, some of them handwritten. Finding a specific document or tracing a specific transaction through that archive manually could take days or weeks.

Veritas Data was engaged to build a platform that would make that archive instantly searchable — so legal counsel could find any document, cross-reference any claim, and surface relevant evidence in seconds rather than days.

The Challenges
  • 60+ years of historical documents across multiple formats — typed, scanned, and handwritten
  • Bilingual archive — documents in both English and Spanish requiring cross-language search
  • Legal research requiring precise, cited answers — not approximate keyword matches
  • Strict security and access control requirements for sensitive litigation materials
  • Need to query across the entire archive in real time during legal proceedings
What Was Needed
  • OCR processing of scanned and handwritten documents into searchable text
  • Bilingual search capability — query in English, find documents in Spanish, and vice versa
  • LLM-powered query interface returning cited, sourced answers
  • Encrypted storage with role-based access controls and full audit logging
  • Cross-document synthesis — connecting evidence across multiple files automatically

A secure, bilingual document intelligence platform on AWS.

We designed and built the platform end-to-end — from document ingestion and OCR processing through the LLM-powered query interface — on AWS, with security and compliance configuration built for sensitive legal materials.

🔍
AWS Textract OCR
Automated extraction of text from scanned PDFs, image files, and legacy documents — including handwritten records — using AWS Textract with custom post-processing for accuracy.
🌐
Bilingual Search (EN/ES)
Cross-language search capability allowing legal counsel to query in English and retrieve relevant documents regardless of their original language — Spanish or English.
💬
LLM Query Interface
Plain-English query interface powered by LLM — ask a legal question and receive a sourced answer with specific document citations, page references, and relevant excerpts.
🔒
Encrypted Secure Storage
All documents stored in encrypted AWS S3 with role-based access controls — only authorized legal personnel can access the archive or query results.
📎
Cross-Document Synthesis
The platform can synthesize evidence across multiple documents automatically — connecting a transaction record to a correspondence to a financial statement in a single query.
📝
Full Audit Logging
Every query, every access, every document retrieval is logged with timestamp and user — providing a complete chain of custody record for litigation use.

Weeks of research. Now minutes.

60+
Years of historical documents — fully indexed and searchable
Minutes
Research time vs. days or weeks of manual document review
EN/ES
Bilingual query capability — search in either language, find documents in both

Legal counsel can now query the entire historical archive in plain English — asking questions like "show me all board resolutions related to foreign deposits between 1955 and 1959" and receiving cited, sourced answers in seconds. Research that previously took days of manual document review now takes minutes, and the platform surfaces cross-document connections that would be nearly impossible to find manually. The complete audit trail ensures every document access is logged for litigation integrity.

Do you have a document library that should be instantly searchable?

Law firms, compliance teams, and operations departments with large document libraries are ideal candidates for this platform. Schedule a discovery call to discuss your archive and what AI-powered search would unlock.