Everpure and Exadel: putting AI to work on the enterprise webEverpure and Exadel: putting AI to work on the enterprise web
Summary
Everpure, the storage and data management platform company formerly known as Pure Storage, uses AI as a working tool on its own web platform.
Two separate pieces of work show what that looks like in practice: an Adobe Experience Manager migration from 6.5 to AEM as a Cloud Service, run across the whole migration lifecycle with Claude, and a new design system whose components are built from Figma with Claude. Exadel's team delivers both with Everpure's product owners and UX experts.
_2026-09-18_14-07-45.png)
Why the work exists
The migration is driven by Adobe's roadmap: support for AEM 6.5 is ending, so the platform has to move to AEM as a Cloud Service. The design system is a separate programme with its own timeline, building the component library Everpure's future web experiences will use. What connects them is the way the team chose to work: apply AI wherever it removes repetitive engineering effort, keep an engineer responsible for every technical decision, and hold the quality bar Everpure's marketing team expects.
This is the first paper in a three-part series. Whitepaper 2 covers Operational Excellence. Whitepaper 3 covers Implementation, Roadmap, and Cloud Deployment.
What the joint team built
The AEM Cloud migration
The team integrated Claude Code with Adobe's MCP server and with the tools already in the development lifecycle, so Claude works at several stages of the migration:
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Content migration
Synchronising content from the AEM 6.5 implementation to AEM Cloud, cutting the manual effort of the move.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Regression testing
through the Jira integration and access to application logs, Claude picks up issues found in post-migration regression testing, analyses failures and proposes causes for the developers to verify.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Deployment
Claude can initiate deployments, and the Slack integration reports deployment activity to the team in the same workflow.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Documentation
Through the Confluence integration, Claude drafts the technical documentation from the work in progress; engineers review and finalise it.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Research
Developers use Claude to explore solutions to migration problems without leaving the development workflow.
Engineers remain responsible for technical decisions, review and validation. The migration runs on this workflow today
The design system.
Everpure's design system lives in Figma. The Exadel front-end team connected Claude to Figma through the Figma MCP, so Claude works across component build, design review and documentation:
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Figma to components
Claude reads the design source and helps turn it into reusable front-end components, which cuts the manual work of interpreting specs, styles and behaviour. The team's estimate so far: this saves about 25 to 50 percent of the development time per component, depending on how complex the component is.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Design consistency
while building components, Claude has flagged discrepancies in the Figma files that the design review had missed, so the team can fix them before they land in the library.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
Storybook documentation
Claude drafts the Storybook pages that document how each component is used; developers review and refine. That keeps the docs in step with the implementation as the library grows.
The front-end team reviews and hardens every component. The design system is a shared foundation, meant for reuse across Everpure websites and applications.
What it means for Everpure
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
The migration runs with a lean AEM team of 2.5 developers. Claude takes the time-consuming tasks across the lifecycle, and the developers spend their time on architecture, complex implementation and engineering decisions.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
The team's estimate to date: the same scope without Claude would have added about two to three months to the migration timeline, or required additional engineering capacity to hold the schedule.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
A design system built from the design source, ready for the web experiences that will consume it.
Alexey Girzhadovich
Chief Enterprise AI and Solutions Officer
A working pattern Everpure's product and UX teams can point to when the next AI question comes up: what it does, what a human still decides, and how quality is checked.
We wanted to see whether AI could do real platform work to our standard. Running the AEM migration with Claude across content, testing, deployment and documentation, under engineering review, and building our design system components from Figma answered that. The work is reviewable, and it is ours.
Susie Hamlin
CTO, EQS Group
Case Studies
You may also be interested in...
