[FULL WORKSHOP] AI Coding For Real Engineers - Matt Pocock, AI Hero (@mattpocockuk )

2026-04-24 · Source: AI Engineer · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, extended

Summary

Matt, an AI teacher, presented a workshop on integrating AI into software engineering workflows, emphasizing that fundamental software engineering principles remain crucial. The core thesis is that while AI is a new paradigm, established practices like small tasks and clear feedback loops are highly effective. The workshop introduced concepts like the "smart zone" and "dumb zone" for LLMs, where performance degrades with increasing context window usage, typically beyond 100K tokens. It also highlighted the "Momento effect," where LLMs forget past context upon clearing. Key tools and techniques demonstrated included the "grill me" skill for achieving shared understanding with an AI agent, generating Product Requirements Documents (PRDs) as destination documents, and structuring tasks using Kanban boards with vertical slices for parallelizable, AFK (Away From Keyboard) implementation. The session also covered the importance of Test-Driven Development (TDD) for AI agents and designing deep modules for improved codebase architecture and testability.

Key takeaway

For AI Engineers and Machine Learning Engineers aiming to integrate AI into their development workflows, focus on establishing clear planning and feedback mechanisms. Your AI agents will perform better when tasks are small, context windows are managed to stay within the "smart zone" (around 100K tokens), and development is structured with vertical slices. Prioritize manual QA and code review to inject human taste and maintain code quality, as AI-generated code, especially in complex front-end scenarios, still requires human oversight to avoid "slop."

Key insights

Software engineering fundamentals, like small tasks and clear feedback, are crucial for effective AI integration.

Principles

• LLMs have a "smart zone" (optimal performance) and a "dumb zone" (degraded performance with large context).
• Prioritize vertical slices over horizontal layers for faster feedback loops in AI-assisted development.
• Deep modules with simple interfaces improve testability and AI's ability to work with a codebase.

Method

The "grill me" skill facilitates shared understanding with an AI by relentlessly interviewing the user. This leads to a PRD, which is then broken into parallelizable, vertically sliced tasks on a Kanban board for AFK agent implementation.

In practice

• Use the "grill me" skill to align with AI on project requirements.
• Break large tasks into small, vertically sliced issues for AI agents.
• Implement TDD with AI to ensure robust code and effective feedback loops.

Topics

• AI Development Workflow
• LLM Context Management
• "Grill Me" Skill
• Product Requirements Document
• Iterative Development Practices

Best for: AI Engineer, Machine Learning Engineer, Software Engineer

Related on AIssential

• Day 4 - Agentic Engineering/VIbe Coding an AI App from Zero To Production | 🔴 Live — AI-assisted development streamlines application building, from backend integration to UI design and testing.
• [D] Self-Promotion Thread — The community is actively developing diverse open-source tools and frameworks for ML education, research, and practical applications.
• BugBash'26 Afternoon of Day 1 — LLMs make code cheap, but core software design, correctness, and reliable system engineering remain human-centric challenges.
• Dispatches from O'Reilly: The accidental orchestrator — Disciplined AI-driven development requires human orchestration of AI agents, utilizing structured practices and critical thinking for effective outcomes.
• The Death of Coding is an Illusion: A Field Guide to the AI Orchestration Era — Effective AI integration requires treating LLMs as orchestratable distributed systems, not magic oracles.
• Closing the knowledge gap with agent skills — Agent skills effectively bridge LLM knowledge gaps for rapidly changing software development practices.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI Engineer.