Skip to content
VirtualOnlinePast event

Decodering: How I Use Claude Code — Multi-Agent Development Workflow

Tools & platforms

Event artwork for Decodering: How I Use Claude Code — Multi-Agent Development Workflow

This event has ended

This listing is kept as an archive. Explore upcoming events to find your next one.

See upcoming events
About this event

What to expect

Online software development meetup hosted by CodeHub Bristol exploring advanced Claude Code workflows beyond basic chat-based assistance. Stephen Leach presents a practical framework for orchestrating AI development agents in a multi-stage pipeline that mirrors real software team collaboration patterns.

What You'll Learn

This session demonstrates how to structure Claude Code as a multi-agent development system handling the complete software lifecycle. You'll see working patterns for:

  • Requirements gathering agents that elicit and document project needs
  • Architecture design workflows where AI assists in system design decisions
  • Implementation pipelines coordinating code generation across modules
  • Automated code review processes catching issues before human review
  • Documentation generation keeping docs synced with implementation

Technical Deep Dive

The talk focuses on three core technical patterns that enable team-scale AI workflows:

Artifact handoff protocols — how custom agents pass context and work products to each other in a structured pipeline rather than ad-hoc prompting

Persistent memory systems — techniques for agents to retain learning and context across sessions, building institutional knowledge over time

Git worktree integration — using git worktrees to enable parallel AI-assisted development streams without session conflicts or branch switching disruption

Stephen shares what actually works in production use, what patterns proved too complex to maintain, and how to adopt these workflows incrementally without disrupting existing development processes.

Who Should Attend

This meetup targets developers and technical leads already using Claude Code or similar AI coding assistants (like Cursor, Windsurf, or GitHub Copilot) who want to level up from single-session prompting to repeatable, team-scale workflows. You'll need familiarity with git version control and command-line tools — no AI or machine learning background required.

The session assumes you've moved past "will AI replace developers?" questions and are focused on practical engineering: how to integrate AI agents into actual development workflows, debug when they fail, and build systems that combine human judgment with AI capability.

Format

Online presentation with live demonstration of the multi-agent

Running an AI event?

List it here, free.

Meetups, workshops and conferences go in front of thousands of builders after a quick review.

List your event