AI Driven Modernization for Financial Services

*---------------------------------------------------------------- IDENTIFICATION DIVISION. PROGRAM-ID. RQVS0042. AUTHOR. SYS. INSTALLATION. [INSTALLATION-ID]. DATE-WRITTEN. [YYDDD]. DATE-COMPILED. SECURITY. PROPRIETARY - [CLIENT-CODE] INTERNAL USE ONLY. *---------------------------------------------------------------- * PGM : RQVS0042 VER : 04.02.11 * SYS : [SYS-ID] ENV : PROD * MOD : [YYMMDD] BY: [UID] REQ : [CHNG-TKT-ID] *---------------------------------------------------------------- * DESC: REQ XTRCT SUMRY / VLDTN RPT - [CLIENT-CODE] [SYS-ABBR] * BATCH OUTPUT TO RQVS0042.RPTOUT.G[NNNN]V[NN] * CALLED BY: RQBATCH1 STEP070 * CALLS : RQVSUT01 RQVSUT02 ERRUTL03 *---------------------------------------------------------------- * CHG LOG: * [YYMMDD] [UID] V04.02.11 GAP RMDTN CNTR FLD ADDED * [YYMMDD] [UID] V04.01.09 SRC REF ACCY CALC CORRECTED * [YYMMDD] [UID] V04.00.00 BASELINE - MIGR FROM RQVS0039 * [YYMMDD] [UID] V03.14.02 ABEND FIX S0C7 ON WS-COV-PCT * [YYMMDD] [UID] V03.14.01 ABEND FIX S0C7 ON WS-TOT-REQS * [YYMMDD] [UID] V03.12.00 ADDED INDP VLDTN SECTION * [YYMMDD] [UID] V03.00.00 INITIAL PROD RELEASE *---------------------------------------------------------------- ENVIRONMENT DIVISION. CONFIGURATION SECTION. SOURCE-COMPUTER. IBM-Z15. OBJECT-COMPUTER. IBM-Z15 MEMORY SIZE 4096 KILOBYTES SEGMENT-LIMIT IS 48. SPECIAL-NAMES. DECIMAL-POINT IS COMMA. INPUT-OUTPUT SECTION. FILE-CONTROL. SELECT RPTOUT ASSIGN TO UT-S-RPTOUT ORGANIZATION IS SEQUENTIAL ACCESS MODE IS SEQUENTIAL RECORD KEY IS WS-RPT-REC-KEY FILE STATUS IS WS-FS-RPTOUT. SELECT CTLTBL ASSIGN TO UT-S-CTLTBL ORGANIZATION IS INDEXED ACCESS MODE IS RANDOM RECORD KEY IS CTL-REC-KEY ALTERNATE RECORD KEY IS CTL-ALT-KEY

## Conclusion The business requirements extraction for the [Client] [System Name] has been successfully completed with [X] total requirements documented across [N] comprehensive requirements documents. The extraction achieved an average coverage of [X]% across all functional domains, demonstrating thorough analysis of the legacy [Technology] application. **Key Strengths of the Extraction:** Comprehensive coverage of all core CRUD operations and business workflows Accurate documentation of [X]+ calculation formulas and [X]+ business rules Complete traceability to source code with file names and line number references Consistent requirement identification with unique IDs per domain Detailed data model specifications including all entities, relationships, and constraints **Areas for Enhancement:** Some edge case and error handling documentation remains at implementation level Minor UI layout details not captured (cosmetic positioning, font sizes) Overall Assessment: The extracted requirements provide a solid foundation for migration planning and implementation. The documentation captures the essential business logic, data structures, integration points, and operational workflows required to build a modern replacement system that maintains full functional parity with the legacy application. The requirements team recommends proceeding with migration planning using these documents as the authoritative source for business requirements. A comprehensive missing functionality audit was conducted against all [N] source code files, identifying and documenting [X] additional requirements that have been incorporated into the relevant requirements documents. **Independent Validation Summary** An independent validation review verified the completeness and accuracy of these requirements. ## Conclusion The business requirements extraction for the [Client] [System Name] has been successfully completed with [X] total requirements documented across [N] comprehensive requirements documents. The extraction achieved an average coverage of [X]% across all functional domains, demonstrating thorough analysis of the legacy [Technology] application. **Key Strengths of the Extraction:** Comprehensive coverage of all core CRUD operations and business workflows Accurate documentation of [X]+ calculation formulas and [X]+ business rules Complete traceability to source code with file names and line number references Consistent requirement identification with unique IDs per domain Detailed data model specifications including all entities, relationships, and constraints **Areas for Enhancement:** Some edge case and error handling documentation remains at implementation level Minor UI layout details not captured (cosmetic positioning, font sizes) Overall Assessment: The extracted requirements provide a solid foundation for migration planning and implementation. The documentation captures the essential business logic, data structures, integration points, and operational workflows required to build a modern replacement system that maintains full functional parity with the legacy application. The requirements team recommends proceeding with migration planning using these documents as the authoritative source for business requirements. A comprehensive missing functionality audit was conducted against all [N] source code files, identifying and documenting [X] additional requirements that have been incorporated into the relevant requirements documents. **Independent Validation Summary** An independent validation review verified the completeness and accuracy of these requirements. **Key Validation Findings:** Total Requirements Verified: [X] across [N] documents ([X] original + [X] from gap remediation) Source Reference Accuracy: [X]% ([N] references spot-checked) Calculation Formula Accuracy: [X]% ([N] formulas verified) Business Rules Coverage: [X]% ([N] of [N] rules documented) Database Objects Coverage: [X]% ([N] tables, [N] stored procedures) Initial Gap Assessment: The validation initially identified [N] potentially undocumented items. A subsequent remediation review determined that [N] of these items ([X]%) are documented using business-focused language rather than internal function names, which is appropriate for business requirements documentation. Revised Coverage Assessment: >[X]% of critical functionality documented following gap remediation. Remaining gaps limited to infrastructure initialization and UI layout details. Validation Status: PASS — Suitable for migration planning and implementation.

Forensic Logic Extraction

PraxentDecode^SM recovers every business rule, compliance validation, integration dependency, and workflow buried in your legacy financial platform.

SME Validation and Restructuring

We auto-generate clarification questions from ambiguous code and validate every rule with your domain experts.

Rebuild on AI-Ready Architecture

Retire your legacy system and rebuild on modern, AI-ready architecture with full governance and regulatory traceability.

Title

Subtitle

Title

Subtitle

Title

Subtitle

Reverse Engineering

Our AI pipelines extract business logic, dependencies, and embedded rules directly from legacy source code:

→ Inventory and Classification
→ Structure Analysis
→ Dependency Analysis
→ Data Model Documentation
→ Business Logic Extraction and SME Question Generation

Specify

Every rule is documented with full traceability. PraxentDecode^SM auto-generates clarification questions from ambiguous logic and routes them to your domain experts for validation.

→ Draft Specification Synthesis
→ Final Specification Assembly

Validate

A multi-agent flow verifies specifications against live legacy system behavior. Every reconciliation rule, every exception classification, confirmed before we move forward.

→ Validate Requirements
→ Cross Check Requirements

Architect

Design the modern system: migration strategy, agentic upgrades, test framework, and governance model mapped to your regulatory environment.

→ Architecture Design
→ Delivery Planning
→ User Stories / Tasks
→ Epics / Releases
→ Dependency / Risk Identification
→ Agentic Upgrade Recommendations

Build and Deploy

We rebuild on AI-ready architecture with parallel validation against the legacy system. When we're done, the legacy platform is retired.

→ Environment and Scaffold Setup
→ Spec-to-Code Generation
→ Automated Testing
→ Spec Tracing
→ Parity Assessment
→ Review and Remediation
→ Deployment and Cutover

CLIENT RESULT

Modernizing an enterprise payment processor with Claude.


    An enterprise payments company needed to modernize its mission-critical reconciliation application, processing 1.5B+ transactions per day.


    The original engineers were gone and there was zero documentation.

PraxentDecode^SMrecovered all the business rules governing OOB classifications, exception ticketing, funds movement logic, and the 7-year retention requirements buried in the code.

A test-harness validated system parity enabling them to rebuild on AI-ready architecture without disruption.

10K

LINES OF CODE
ANALYZED

700+

BUSINESS REQUIREMENTS  RECOVERED

125+

SME CLARIFICATION  QUESTIONS GENERATED

SECURITY vulnerabilities SURFACED

Get results like these →

Praxent

Decode

AI-Accelerated Legacy Modernization We extract and validate every business rule in your legacy platform, then retire it and rebuild on AI-ready architecture.

[You are here]

Praxent

Activate

AI-Native Financial Products We rebuild on modern architecture with AI at the center. Governed, auditable, production-grade.

Learn about Activate →

Praxent

Redesign

Intelligent Financial Experiences We design the intelligent, conversational, and personalized interfaces your customers and teams interact with every day.

Learn about Redesign →

‹ ›

Swipe to see more

Paul Tidwell VP, Technology and AI, Praxent

With most modernizations, teams begin building before they actually know what the old system does. The first thing we do is understand exactly what your system does. We spend real time reading the legacy code, mapping every dependency, surfacing every assumption, and pulling out every business rule.

Only then do we build against that. AI does the volume work while our engineers make the calls that require judgment. That combination is how you build something modern without losing what matters.

Meet Our AI Modernization Experts

Kevin Hurwitz

Managing Partner

With 25+ years guiding 500+ organizations, Kevin brings his expertise to help financial leaders apply AI across product, engineering, and operations to drive measurable, real-world results.

Paul Tidwell

VP of Technology and AI

Paul designs enterprise AI strategies that turn legacy cores into secure, production-ready AI ecosystems. His 20+ years of experience drives AI-native SDLC and platform modernization efforts built for scale, security, and long-term impact.

Ben Bays

VP of Enterprise Solutions

Ben applies AI in complex, regulated environments to reduce tech debt, strengthen DevSecOps, and modernize software delivery. With 20 years in fintech, he connects data, cloud, and engineering to embed AI into enterprise workflows.

Matthew Wooley

Technical Director

Joseph Young

Technical Director

Danial Maloney

Technical Director

Mauricio Araya

Technical Director

Rodrigo Callirgos

Principal Software Engineer

Diego Romero

Principal Software Engineer

Marcelo Dornelles

Senior Software Engineer

Nilo Arnt

DevOps Engineer

Find out what's inside your platform.

Paul-AI-Modernization-Experts-Praxent (1)

Kevin — AI Modernization Expert at Praxent

START WITH CLARITY

Get a Platform Intelligence Report

We map business logic density, undocumented rules, compliance-critical workflows, integration dependencies, and modernization priorities inside your current system.

It's the first step toward AI Command.

SPECIALIZED AI EXPERTISE

The Praxent Difference.

Our commitment to quality goes far beyond technology. We’re not a global consulting firm. We’re a boutique team of experts in financial technology with proven expertise in AI solutions.

× No one-size-fits-all frameworks × No two-year contracts × No mass out-sourcing to teams outside your time zones. Just experienced professionals dedicated to solving real business problems. Think of us as part of your team. Everyone else does.

Let's Talk →

AI-DRIVEN MODERNIZATION

Praxent Decode SM

Talk to our AI experts ↓