How Marks & Spencer uses GitHub Copilot to scale retail engineering
GitHub presents a look at the engineering workflows supporting M&S’s food, fashion, and beauty operations, showing where AI agents assist delivery and where developers remain responsible for decisions.
AI-assisted delivery at retail scale
Marks & Spencer operates more than 1,000 stores across 70 global markets. As the business scales, its engineering teams use GitHub Copilot and agentic workflows across the software development lifecycle.
The workflows use specialized AI agents for three areas:
- Ticket scaffolding: automating the scaffolding work associated with development tickets.
- Code generation: assisting engineers with producing code.
- Pull request reviews: automating review work as part of the delivery process.
M&S reports productivity uplifts of up to 75% for key tasks. That figure applies to selected tasks rather than representing a stated improvement across all engineering work.
Developers retain architectural control
The approach combines automated assistance with developer ownership. While agents support ticket preparation, implementation, and review, engineers remain in control of architectural decisions.