Skip to main content
Version: 0.1.123

Features

Executive Summary​

Repave.ai is an AI-powered legacy application modernization platform that helps enterprises accelerate digital transformation by automatically analyzing, understanding, and modernizing legacy codebases. Our platform reduces modernization costs by up to 70% and accelerates time-to-market by 10x compared to traditional manual rewrite approaches.

Core Capabilities​

1. Intelligent Legacy Analysis​

Automated Codebase Understanding

Repave.ai uses advanced AI to automatically analyze and understand your legacy applications through a comprehensive 5-step analysis pipeline:

  • Automatic File Discovery - Scans and catalogs all source files across your codebase
  • Dependency Mapping - Builds complete dependency trees showing how modules interconnect
  • Database Access Analysis - Identifies all database interactions and data access patterns
  • User Interface Discovery - Automatically detects and catalogs all UI components and screens
  • User Action Extraction - Identifies business workflows and user interactions
  • Multi-language support - Works across Java, C#, VB.NET, PHP, Delphi/Pascal, Python, JavaScript/TypeScript, and other legacy stacks

Business Value:

  • Zero manual documentation required
  • Complete visibility into legacy application structure in hours, not months
  • Uncover hidden dependencies that cause 80% of migration failures

2. Knowledge Graph Technology​

Deep Codebase Intelligence

Our proprietary knowledge graph engine creates a comprehensive model of your application, capturing:

  • File-to-file dependencies and import relationships
  • Database table access patterns and CRUD operations
  • UI screen flows and navigation paths
  • User actions and business workflows
  • Feature-to-code traceability with line-level precision
  • Source File Tracing - jump from any Gherkin scenario directly to the exact legacy files, routines, and line ranges that implement it — bidirectional navigation between spec and source
  • End-to-end calling chains — surfaces hidden coupling, dead code, and blast radius before you touch anything
  • Static and dynamic coverage analysis that identifies uncovered code segments, with AI-powered feature generation to close coverage gaps

Business Value:

  • Instant answers to "what will break if we change X?"
  • Eliminate 90% of unexpected migration issues
  • Enable confident refactoring decisions with complete impact analysis

3. AI-Powered Architecture Visualization​

Automatic Architecture Documentation

Repave.ai generates industry-standard C4 architecture diagrams and entity-relationship diagrams automatically as standalone HTML documents with inline SVG diagrams:

  • C1 (System Context) - High-level system boundaries and external dependencies
  • C2 (Container) - Application containers, services, and major components
  • C3 (Component) - Internal component structure and interactions
  • C4 (Code Level) - Class diagrams and detailed code relationships
  • Database ER Diagrams - Complete entity-relationship visualization

Business Value:

  • Instant architecture documentation for legacy systems with zero documentation
  • Accelerate new team member onboarding by 80%
  • Enable better architectural decisions with visual insights

4. Behavior-Driven Development Methodology​

Automated Feature Specification & Implementation

Transform legacy code into modern, well-tested features using industry-standard BDD practices:

  • Automatic Gherkin Generation - AI reverse-engineers legacy code into executable business specifications, capturing business logic without relying on tribal knowledge or outdated documentation
  • Single Source of Truth - Specifications are readable by business stakeholders and executable by engineers
  • Gherkin-to-Legacy Mapping - Every specification links back to the exact source routines it describes — full traceability from business rule to implementation
  • AI-Powered Implementation - Orchestrates the full feature lifecycle: worktree setup → AI code generation → unit testing → BDD testing → regression testing → git merge
  • Conversational Refinement - Implementations, prototypes, and API specs are iteratively improved through chat-based feedback loops, not one-shot generation
  • Comprehensive Testing - BDD scenarios with Cucumber integration ensure correctness
  • Regression Protection - Automated regression testing prevents breaking existing functionality

Business Value:

  • Bridge business and technical teams with human-readable specifications
  • Achieve 95%+ test coverage automatically
  • Reduce post-migration defects by 85%

5. Characterisation Testing​

Prove Behavioural Equivalence Between Legacy and Modern Systems

  • Auto-generates characterisation test suites from Gherkin specs, proving behavioural equivalence between legacy and modernised systems
  • Doubles as your regression suite - the same characterisation tests that prove parity on day one become your permanent regression safety net, running on every implementation pass and every code change
  • Test-first migration - every scenario is green before code ships — no silent regressions, no "it worked before" surprises
  • Supports Docker-based and Bring-Your-Own-Scripts (BYOS) modes for running tests against live legacy applications
  • False positive detection validates that passing tests reflect genuine behavioural equivalence, not superficial stubs

Business Value:

  • Ship modernised code with mathematical confidence in behavioural parity
  • Catch regressions before they reach users
  • Eliminate "it worked in legacy" debates with executable evidence

6. Test Replay​

Quick Verification Without Re-Running the Full Suite

  • Replay any past test execution from its recorded trace — see exactly what happened, step by step, without re-running the test
  • Inspect captured inputs, outputs, intermediate state, and assertion results for every step of a scenario
  • Compare runs side-by-side to spot when and where behaviour diverged between legacy and modernised implementations
  • Enables fast triage: stakeholders and engineers can verify a fix in seconds instead of waiting for a full regression cycle

Business Value:

  • Dramatically tighten the feedback loop during implementation and debugging
  • Give reviewers a lightweight way to audit test results without infrastructure access
  • Preserve a permanent, inspectable record of every test run for compliance

7. Personas Extraction​

Understand Who Uses the System

  • Automatically extracts user personas from legacy code, UI flows, and authentication/authorisation patterns
  • Maps each persona to the features, screens, and workflows they exercise
  • Surfaces role-based access boundaries implicit in legacy code — so modernised apps can implement them explicitly
  • Feeds persona context into UI prototype generation and scenario sequencing for persona-aware testing

Business Value:

  • Replace tribal knowledge about "who does what" with explicit, traceable persona definitions
  • Ensure modernised systems preserve legacy role semantics
  • Enable persona-driven UX reviews with stakeholders

8. External System Mapping & Test Strategy​

Identify and Isolate Third-Party Dependencies

  • Auto-discovers external systems the legacy application depends on: third-party APIs, message queues, file drops, email gateways, payment providers, mainframes, etc.
  • Classifies each dependency and uses the Mock test strategy:
    • lightweight in-memory doubles for fast feedback cycles
    • local service stubs such as WireMock, MailHog, or MinIO when the app needs a real endpoint shape
  • Generates the corresponding test doubles, fixtures, and integration harnesses automatically
  • Keeps external-system tests isolated — no accidental reliance on live services in CI

Business Value:

  • Eliminate flaky tests caused by external dependencies
  • Make integration risk visible and intentional, not accidental
  • Accelerate CI/CD by picking the right test double for each dependency

9. Scenario Intelligence​

Coordinate Implementation Across Features

  • Scenario sequencing determines optimal implementation order based on logical dependencies across features
  • Scenario triage detects which specifications need regeneration after upstream code changes
  • Scenario comparison performs semantic diffs across features to identify overlaps and conflicts
  • Gap analysis finds orphaned files not covered by any feature and recommends new specifications to close coverage

Business Value:

  • Avoid rework from implementing features in the wrong order
  • Keep specifications in sync with evolving code automatically
  • Guarantee full feature coverage before cutover

10. UI Prototype Generation​

From Specs to Interactive Prototypes

  • Generates interactive HTML prototypes directly from Gherkin scenarios and legacy screen analysis
  • UI Component Picker - click any element in the live prototype to select it, then request targeted refinements through conversational feedback — a tight design feedback loop with no ambiguity about which component you meant
  • Iterative refinement through conversational feedback — stakeholders can review and shape UX before production code is written
  • Version tracking preserves the full design evolution history

Business Value:

  • Validate UX with stakeholders before engineering investment
  • Catch design issues when they're cheap to fix
  • Maintain a complete audit trail of design decisions

11. API-First Modernisation​

Contract-First APIs from Legacy Services

  • Auto-generates OpenAPI 3.0 specifications from Gherkin scenarios and legacy service interfaces
  • New APIs are contract-first: consumers can integrate in parallel with backend development
  • Built-in OpenAPI validation and an API playground (Swagger UI) for interactive testing
  • Iterative spec refinement through chat-based feedback

Business Value:

  • Unblock frontend and consumer teams earlier in the modernisation cycle
  • Enforce API contracts before implementation
  • Accelerate integration testing with Swagger UI out of the box

12. Modernised Application Generation​

Full-Stack Web Apps & Batch Jobs, Generated End-to-End

Repave.ai doesn't just convert code — it generates complete, production-ready applications from the specs it extracts:

Full-Stack Web Application Generation

  • Generates a complete modernised application from extracted features: backend services, REST APIs, database schema, and frontend UI — all wired together
  • Modern stack by default (Next.js / React / TypeScript / Prisma), with architectural patterns aligned to microservices or cloud-native targets
  • Business logic preserved from the legacy source; technology stack fully refreshed
  • Generated apps ship with their BDD test suite, CI configuration, and seed data

Batch Job Application Generation

  • Generates complete batch job applications targeting modern databases (MySQL, PostgreSQL, cloud-native)
  • Legacy Informix / DB2 / Oracle scripts are converted with schema, stored procedures, and data transformations regenerated for the target platform
  • Generated batch apps ship with their converted test suite and baseline validation harness
  • Preserves processing guarantees and data integrity across the migration

Business Value:

  • Single platform for full-stack web app and batch job modernisation
  • Reduce vendor lock-in for legacy databases
  • Generated apps are not scaffolds — they're functional, tested, and ready to deploy

13. Database Migration Excellence​

Legacy-to-Modern Database Conversion

Seamlessly migrate from legacy databases to modern platforms:

Supported Source Databases:

  • IBM Informix
  • IBM DB2
  • Oracle Database
  • Other legacy RDBMS

Supported Target Databases:

  • MySQL
  • PostgreSQL
  • Cloud-native databases (AWS RDS, Azure SQL, etc.)

Migration Capabilities:

  • Schema extraction and AI-powered type-mapping conversion with conversion reports flagging items that need manual review
  • Query translation and optimization
  • Stored procedure modernization
  • Data type mapping and validation
  • Automated test suite migration - baseline validation against the legacy system, then test conversion and re-validation on the target
  • Container-based execution environments for both source and target database stacks

Business Value:

  • Eliminate expensive legacy database licenses (save $100K-$1M+ annually)
  • Enable cloud migration for cost savings and scalability
  • Reduce database maintenance costs by 60%

14. Safe & Isolated Implementation​

Git Worktree Technology & Incremental Migration

Repave.ai uses advanced git worktree isolation to ensure safe feature implementation:

  • Isolated Environments - Each feature implemented in dedicated git worktree
  • Parallel Development - Multiple features can be worked on simultaneously
  • Dual-Branch Architecture - Legacy and modernised code live side by side; feature work never touches the legacy branch until validated
  • Automatic Rebase - Keeps implementations up-to-date with the Repave integration branch (repave by default)
  • Regression Testing - Validates no existing functionality is broken
  • Automated Merging - Safely integrates completed features, with optional push to the remote repository
  • Reversible by Design - Rollback is always an option at every phase
  • Implementation tracking is evidence-backed - Per-scenario implementation state is written by the implementation pipeline as the agent completes each scenario. A feature is only marked complete from validated test evidence: /repave:upload-test-result in the Repave web IDE runs repave features submit-tests against the recorded run output — found wherever the checkout's runner writes it, which is the shared test folder in a combined workspace — refuses to submit if anything changed since that run, and the server flips the feature to completed only when every scenario is green. Each scenario's status on the feature page is its most recent result, whichever produced it — an upload replaces a failure an earlier implementation job left, and a later job run replaces the upload. There is no manual override from the web UI.

Business Value:

  • Zero risk to production code during modernization
  • Continue normal development while modernization progresses
  • Accelerate modernization with parallel feature implementation
  • Incremental cutover at the feature level — no big-bang migrations

15. Real-Time Progress Monitoring​

Complete Visibility & Control

Monitor every aspect of your modernization project in real-time:

  • Job Progress Streaming - Server-Sent Events (SSE) for live progress updates on all long-running operations
  • Pipeline Status Dashboard - Visual representation of analysis pipeline stages
  • Feature Implementation Tracking - Monitor BDD test execution and results
  • AI Session Logs - Full transparency into AI decision-making process
  • Metrics & Analytics - AI cost tracking and session metrics — full visibility into token usage and spend per project

Business Value:

  • Full transparency eliminates "black box" concerns
  • Proactive issue detection and resolution
  • Accurate project timeline forecasting

16. Enterprise-Grade Scalability​

Asynchronous Job Processing

Built for enterprise-scale modernization projects:

  • 70+ Job Types - Parallel execution, cancellation, and timeout/resume across the full analysis and implementation pipeline
  • Concurrent Job Processing - Configure worker count based on infrastructure
  • Job Cancellation & Recovery - Gracefully handle interruptions and failures
  • Resource Management - Automatic throttling and load balancing
  • Database Optimization - Efficient handling of large codebases (10K+ files)
  • Multi-Tenant Ready - Project-level access control and internal API authentication

Business Value:

  • Modernize large enterprise applications (500K+ LOC)
  • Predictable resource utilization and costs
  • Handle multiple projects simultaneously

17. Cloud-Native Deployment​

Kubernetes-Ready Platform

  • Kubernetes-native agent execution with Terraform-managed infrastructure, RBAC, and namespace isolation
  • Containerised agent runners (Dockerfile.agent-runner, Dockerfile.app) for consistent execution environments
  • GCS-backed storage for agent artefacts and session data
  • Deployable to any container platform — no cloud vendor lock-in by default

Business Value:

  • Deploy where your compliance and data residency rules require
  • Scale horizontally with your existing cloud infrastructure
  • Avoid vendor lock-in while benefiting from managed cloud services

Key Differentiators​

vs. Manual Rewrite​

  • 70% cost reduction - Automate analysis and implementation that typically takes months
  • 10x faster - Complete modernization in weeks vs. years
  • Higher quality - Automated testing ensures correctness

vs. Other Tools​

  • Complete Solution - Analysis + Implementation + Testing in one platform
  • BDD Methodology - Business-readable specifications bridge stakeholder gaps
  • Knowledge Graph - Deeper understanding than simple static analysis tools

vs. Offshore Teams​

  • Consistent Quality - AI doesn't have bad days or skill gaps
  • No Knowledge Transfer - AI learns your codebase instantly
  • 24/7 Operation - Continuous progress without timezone delays
  • Full Traceability - Complete audit trail of all decisions and changes

Use Cases​

Financial Services​

"Modernize mainframe COBOL applications to cloud-native microservices while maintaining regulatory compliance and audit trails"

Healthcare​

"Convert legacy patient management systems to HIPAA-compliant modern architectures with zero downtime"

Retail & E-Commerce​

"Upgrade monolithic e-commerce platforms to scalable microservices while preserving complex business rules"

Manufacturing​

"Migrate legacy ERP batch jobs from Informix to cloud databases, reducing operational costs by 60%"

Government​

"Modernize legacy citizen services applications with full documentation for compliance and transparency"

Technical Specifications​

Technology Stack​

  • AI Engine: Anthropic Claude (Sonnet 4.5)
  • Frontend: Next.js 15, React 19, TypeScript, Tailwind CSS
  • Backend: Next.js API routes, Prisma ORM
  • Database: SQLite, PostgreSQL (configurable)
  • Graph Database: Neo4j for knowledge graph
  • Testing: Cucumber.js, Jest
  • Version Control: Git with worktree support

Deployment Options​

  • Cloud SaaS - Fully managed platform hosted by Repave.ai
  • Private Cloud - Dedicated instance in your cloud environment
  • On-Premises - Self-hosted for maximum data control

Security & Compliance​

  • Data Privacy - Your code never leaves your infrastructure (on-prem/private cloud)
  • Access Control - Role-based permissions and project isolation
  • Audit Logs - Complete traceability of all AI actions and code changes
  • Encryption - Data encrypted at rest and in transit
  • Compliance Ready - SOC 2, GDPR, HIPAA-ready architecture

Getting Started​

1. Upload Your Legacy Code​

Simply upload a ZIP file of your legacy codebase (or connect to your repository)

2. Automated Analysis​

Our AI analyzes your application in hours, generating:

  • Architecture diagrams
  • Database schemas
  • Feature specifications
  • Dependency graphs

3. Review & Prioritize​

Review discovered features and prioritize modernization order based on business value

4. AI Implementation​

Our AI implements features using modern best practices with comprehensive tests

5. Validate & Deploy​

Review implementations, run tests, and deploy to production with confidence

Success Metrics​

Typical Customer Results:

  • Analysis Time: 2-4 hours for 100K LOC (vs. 2-3 months manual)
  • Cost Reduction: 70% lower than manual rewrite or offshore teams
  • Time to Market: 10x faster implementation
  • Test Coverage: 95%+ with automated BDD scenarios
  • Defect Rate: 85% reduction in post-migration defects
  • Documentation: 100% code-to-specification traceability

Pricing​

Contact our sales team for custom enterprise pricing based on:

  • Codebase size (lines of code)
  • Number of applications
  • Deployment model (SaaS vs. on-premises)
  • Support level required

Start with a Free Pilot: Analyze one application (up to 50K LOC) free to see the value firsthand.

Support & Services​

Professional Services​

  • Migration Strategy Consulting - Expert guidance on modernization approach
  • Proof of Concept - Validate Repave.ai on your specific codebase
  • Custom Integration - Connect to your existing tools and workflows
  • Training & Onboarding - Get your team up to speed quickly

Support Tiers​

  • Standard: Email support, 24-hour response time
  • Premium: Priority support, 4-hour response time, dedicated Slack channel
  • Enterprise: 24/7 support, dedicated success manager, custom SLAs

About Repave.ai​

Repave.ai is on a mission to help enterprises unblock innovation by modernizing legacy applications that hold them back. Our AI-powered platform makes legacy modernization affordable, fast, and reliable - enabling companies to focus on innovation rather than maintaining technical debt.

Contact Us:

Last Updated: April 2026