Devono

Devono is a London office-space tenant advisor helping businesses find, negotiate and secure the right workspace.

Their existing WordPress website had grown into a large content and property platform, but the technology behind it was becoming increasingly difficult to scale. Thousands of property listings, images, articles and location pages had to be maintained, while the team also needed better control over content, search and incoming enquiries.

The rebuild came with an additional challenge: Devono could not afford to lose the visibility and content accumulated over years of operating the existing platform. Existing URLs needed to remain accessible, large amounts of WordPress data had to be migrated safely, and the new experience had to work smoothly across desktop and mobile.

At the same time, Devono wanted the new platform to do more than simply display office listings. They wanted visitors to be able to interact with the website more naturally, discover relevant properties faster and turn those conversations into qualified enquiries.

Interactivated rebuilt Devono.com as a modern Next.js platform, using Claude Code extensively throughout the development process.

The new system combines the public website, property search, content management and internal administration in one scalable application. Supabase powers the underlying database, authentication and storage, while the custom CMS gives Devono's team direct control over listings, articles, area guides, SEO settings, enquiries and other day-to-day content.

A major part of the project was the migration from WordPress. Rather than moving content manually, we created a repeatable migration pipeline that transferred 1,295 property listings, around 9,400 images, 275 articles and 19 area guides into the new platform.

The website also introduces a much richer property discovery experience. Visitors can search by area, street, postcode or Tube station, while the architecture supports dedicated area pages, service content and a large number of database-driven URLs without making the platform difficult to maintain.

Claude Code played a central role in implementation. Our developer used it to accelerate development, translate approved designs into production components, work through the existing codebase and support testing throughout the project. Human review remained part of every stage: architecture, implementation decisions, testing and deployment stayed under developer control.

The new Devono platform also includes a set of AI-powered property advisors built with the Claude API.

Instead of a traditional scripted chatbot, the advisors can respond to visitors based on the context of the conversation and the property they are viewing.

A visitor can explain what they are looking for, and the advisor can help narrow down suitable office options, retrieve listing information and continue the conversation naturally.

When a visitor is ready to speak with Devono, the system can capture their contact details and requirements and pass the lead directly to the team.

The same underlying AI engine supports several advisor personas across the website. Devono can manage those advisors from the CMS, including their profile, behaviour and availability, without requiring a development change every time.

The system also includes safeguards designed for a production environment. The advisors work only with available property information, are instructed not to invent prices or buildings, and clearly indicate that responses are AI-generated.

The result is a complete replacement for Devono's previous WordPress setup, built around a modern architecture that can support both the current business and future growth.

The new platform brings property search, content, SEO, administration and lead generation into one connected system.

Devono now has:

• a modern, responsive property platform built in Next.js;
• a custom CMS covering listings, content, enquiries and SEO;
• 1,295 migrated property listings and approximately 9,400 migrated images;
• 275 migrated articles and 19 area guides;
• advanced location-based office search;
• AI-powered advisors connected directly to live property data;
• automated lead capture from AI conversations;
• a scalable content and SEO structure supporting around 1,600 database-driven URLs;
• a platform the internal team can manage without relying on developers for everyday content changes.

The project also demonstrates how AI-assisted software development can be used in a real production build. Claude Code helped accelerate implementation, while the Claude API became part of the final customer-facing product itself.

Instead of adding AI as a standalone feature, it became part of both how the platform was built and how Devono now serves visitors on the website.