Much has changed in the fields of LLMs and AI since we announced our semantic search capability through integration with third-party providers (Hugging Face, txtai) in May 2023. The technologies have improved, and today AI-driven features surround almost every service and product we interact with daily.

​​​​​​​

Since the initial release, we have added support for more providers (e.g. OpenAI) and extended the coverage to more asset types, like Documents (Liferay DXP 2024.Q4+) or Custom Objects and Modifiable System Objects (coming in Liferay DXP 2025.Q4).

One thing has not changed: regardless of whether you just want to improve the overall relevance of the search results of your Liferay-based solutions or you are also aiming to create RAG experiences, you need good context as input. Good context starts with capturing the meaning of your content and the end-user's searches.

Register #LRDEVCON25 and join my session, where I'll showcase another way to create semantic search experiences through configuration by leveraging built-in models (ELSER, E5) and other generally available Elastic Stack capabilities that you can unlock with your Liferay Enterprise Search (LES) subscription.