AI accelerated legacy modernisation
summary
Claims AI-driven transpilation and automation cut migration effort roughly 4x speedup in practice by iterating prompts and agents; concrete pipeline builds a JSON dependency map from LoopBack code, loads it into Neo4j dependency graph and uses VS Code plugins plus LLMs (ChatGPT and Claude) to generate target slices (Spring Boot + Angular or native mobile). Per-slice times drop from 8 → 5 → 3 → ~1.5 days per slice after prompt tuning. Works but needs human review, equivalence tests, and careful context-window management.
AI-written summary · about
chapters (8)
- 0:00Introduction and legacy risks
- 4:54AI potential for modernization
- 7:38Case study and AI approach
- 15:49Migration speed and results
- 16:09Four modernization blockers
- 20:46Building a modernization factory
- 23:14Lean tech principles
- 27:32Q&A session
related talks
28mAI Migration at SCALE: Make Complex Code Tasks 10X Faster (The Blueprint)
39mGenAI-assisted reverse engineering for incremental Legacy modernisation
30mAI First Legacy Migration
26mAI and APIs: Ensuring Platform Migration Reliability with AI
33mAI assisted legacy modernisation : From UX to code
37mAI-Driven Developer Productivity at Scale
28mApplication modernisation with generative AI
26mA Pragmatic Journey from Monolith to Microservices Powered by AI:GenAI