Legacy Code with AI: Understand, Modernize, Migrate

Modalità
Online
Lingua
en
Livello
practitioner

Il corso

Most enterprise engineering is brownfield. Agents finally made it tractable: comprehension, safety nets, strangler slices, and migrations, practiced on a provided legacy codebase, applied to yours.

Identità del corso

Livello

practitioner

Lingua

en

Programma e obiettivi

Obiettivi
  • - Turn a system nobody fully understands into a verified map you can defend in a design review.
  • - Pin the current behavior of untested code with agent-generated golden-master tests, so refactoring finally has a tripwire.
  • - Carve one already-tested seam out of the monolith with agent help, then run it safely behind the code it replaces.
  • - Run the deterministic recipe first, then let an agent clean up only what it leaves behind.
  • - Translate one already-migrated module into a second language, then ship it with the same proof that showed the original worked.
  • - Generate the README, the ADRs, and the CI check the codebase never had, then prove they stay honest as the code keeps moving.
Programma
  • Url: https://aiacademy.anthropos.work/chapters/legacy-code-ai-intro/ · Legacy Code with AI: Start Here · Position: 1 · A 10-minute orientation to the Legacy Code with AI path: why brownfield engineering finally has a safe agent loop, how the seven chapters build on each other, and where to start.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-map-codebase/ · Map an Unknown Codebase · Position: 2 · Turn a system nobody fully understands into a verified map you can defend in a design review.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-characterization-tests/ · Characterization Tests · Position: 3 · Pin the current behavior of untested code with agent-generated golden-master tests, so refactoring finally has a tripwire.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-strangler-pattern/ · The Strangler Pattern · Position: 4 · Carve one already-tested seam out of the monolith with agent help, then run it safely behind the code it replaces.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-framework-migration/ · Migrating to Java 21 · Position: 5 · Run the deterministic recipe first, then let an agent clean up only what it leaves behind.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-language-translation/ · Language Ports, Verified · Position: 6 · Translate one already-migrated module into a second language, then ship it with the same proof that showed the original worked.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-doc-recovery/ · Documentation, Recovered · Position: 7 · Generate the README, the ADRs, and the CI check the codebase never had, then prove they stay honest as the code keeps moving.
  • Url: https://aiacademy.anthropos.work/chapters/legacy-rewrite-refactor-memo/ · Rewrite or Refactor · Position: 8 · Turn agent-gathered evidence about cost, risk, and reversibility into a decision memo executives can actually hold you to.

Edizioni

Edizioni

Course Mode: online · Course Workload: PT160M · Mode: online

Corsi simili

  • Deep Learning

    A decision-framework deep learning course for engineers. Choose PyTorch vs TensorFlow, judge depth vs classical ML, weigh transfer learning, and reason about CNNs. 7 chapters.

    Prezzo su richiestaScopri
  • AI Foundations

    Use ChatGPT, Claude, and Gemini with confidence at work: learn the vocabulary, how models work, when to verify them, and reusable prompts. 8 chapters, foundations level.

    Prezzo su richiestaScopri
  • AI Engineering Foundations

    Ship AI features to production: prompting, RAG, structured outputs, fine-tuning, and inference tuning. Hands-on, free, 12 chapters (~4.3h) for engineers.

    Prezzo su richiestaScopri
  • Power Platform AI

    Ship AI inside Microsoft Power Platform: AI Builder, Power Apps and Power BI Copilot, Dataverse agents, plus DLP governance. 6 chapters, ~2h, practitioner level.

    Prezzo su richiestaScopri
  • AI Coding for Beginners

    Build real software with AI without writing code: custom assistants, AI agents, dashboards, and clickable prototypes. Hands-on with Claude Code, ChatGPT, and Lovable. 7 chapters…

    Prezzo su richiestaScopri
  • What Are AI Agents?

    Understand how tools, memory, and goals turn a chatbot into an AI agent that does work, why agents fail, and how to direct them. No code. 6 chapters, ~95 min, no experience needed.

    Prezzo su richiestaScopri
  • On-Device & Edge AI

    Run AI directly on a phone or Mac with no cloud round-trip. Build with Apple Foundation Models, Gemini Nano, and MLX across 4 advanced chapters for app engineers.

    Prezzo su richiestaScopri
  • Copilot Studio

    Build and ship custom AI agents in Microsoft Copilot Studio: topics, RAG knowledge sources, connectors, actions, and DLP governance. 4 chapters, practitioner level.

    Prezzo su richiestaScopri