Design partner story · Retail
Keeping customer and payment data out of AI agents’ reach at global scale.
A live design partnership focused on real developer workflows, real enterprise systems and measurable control.
IndustryRetail and e-commerce
Agents in useClaude Code, Gemini CLI and other tools
StatusDesign partnership active
The challenge
The risk arrived before the control.
With thousands of developers and hundreds of internal systems, AI agent use had spread faster than any policy could follow. Agents could touch customer records, payment flows and supplier systems. Card data is under strict compliance rules, so a single agent pasting the wrong data into the wrong tool was a real risk.
What they’re doing with RayShield
- Discovering which AI tools are in use across teams, including ones security didn’t know about
- Enforcing role-based access so different teams’ agents reach only their own systems
- Using built-in data loss prevention to stop card and customer data from leaving before it’s sent
- Controlling which connectors—Slack, Jira and email—agents are allowed to use
What success looks like
- Sensitive data blocked from leaving through agents, on every run
- A clear view of AI usage for every team from one console
- Connector requests approved by security, not self-served