Marc Hermann and Nataliia Baldina are representing Portalworks at Liferay DEVCON at the QEII Centre in London this week. The main programme runs on 3 and 4 November, followed on 5 November by an Unconference whose topics are set by the attendees on site.
The conference forum lists 32 talks, and judging by their titles almost half deal with AI and agents. We went through the programme looking for the topics that matter most for running Liferay projects.
SEO & AEO Studio: visibility in search and AI answers
For marketing and editorial teams, the SEO & AEO Studio is the most important new product. Liferay announced the beta on 2 November, with general availability planned for the first quarter of 2027. AEO stands for answer engine optimization, meaning visibility in the answers given by AI assistants.
- The studio continuously checks sites for issues with metadata, indexing and crawling, and for AI search readiness.
- It shows which AI crawlers such as GPTBot, ClaudeBot or PerplexityBot visit a site and whether AI assistants cite it as a source.
- The data comes from Google Search Console, PageSpeed Insights and Liferay Data Platform.
- AI agents draft fixes such as meta descriptions, titles or FAQ sections. Changes go through the existing versioning and approval workflows of Liferay DXP.
How much the studio helps in day-to-day projects can only be judged on real sites. What is already included in Liferay 2026.Q3 is covered in our article on the SEO package.
AI Hub and MCP: agents with access to Liferay content
Several talks cover Liferay AI Hub, including enterprise RAG and the coordination of multiple agents. AI Hub is a low-code SaaS product for building agents based on language models and MCP servers. The public beta started in June 2026, and agents with multiple data sources are announced for 2026.Q4.
Another talk shows how to build an MCP server for Liferay data. With Liferay’s MCP server, agents inherit the permissions of the respective user and only see content that user can access. Read more in our article on the MCP server in 2026.Q3.
Operations and upgrades: fewer image builds, more GitOps
“Build Liferay without Building Liferay” presents Overlay, a new way of customizing Liferay without rebuilding Docker images. It is based on Liferay CNE (Cloud Native Experience), Kubernetes, Helm and ArgoCD. “GitOps Your Liferay” goes from zero to a running cluster.
“Making Liferay Upgrades Lighter with AI” looks at the use of AI in upgrades. This affects us directly, as upgrades and migrations are part of our daily work.
Privacy, accessibility and search
- Consent: “Building Trust” looks at when Liferay should manage consent itself and when a consent management platform should take over. The session resources include mapping examples for Cookiebot, OneTrust and Usercentrics, among others.
- Accessibility: “AI for Accessibility” and “From Red to Green” deal with improving accessibility, the latter together with SEO.
- Search: “How Good Is Hybrid Search, Really?” asks how good hybrid search actually is.
What comes next
On 18 November, Liferay will present the 2026.Q4 release in a webinar. Announced highlights are AI Hub Agents with multiple data sources, AI-assisted content optimization and the SEO & AEO Studio.
Are you in London too? Feel free to talk to Marc and Nataliia on site. If you would like to know what is relevant for your portal, get in touch.
