Hyland Community
Help
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

The Open Source Cloud Content Repository: What's Next After CommunityLIVE US

angelborroy
Community Manager Community Manager
Community Manager
8 hours ago

A roadmap update for developers, partners, and customers

At CommunityLIVE US 2026 (May 30–June 2, Kissimmee), we introduced the Hyland AI Ready Hub and demonstrated permission-aware semantic search over Alfresco and Nuxeo content using hxpr and the Content Lake architecture. The session “Exploring the Hyland AI Ready Hub: An Open-Source Approach to Data AI Readiness” was delivered by Bertrand Chauvin (Lead Product Manager, Hyland AI), and the technical preview was published on this blog on June 1.

The response has been strong. Developers, partners, and customers (especially across EMEA) are asking how this fits into the on-premises and self-managed AI story for existing Alfresco deployments, and when they can start working with it.

This post answers that: what’s coming, when, and how to get involved today.

Where Content Lake is today

What we showed at CommunityLIVE US was a preview, a working demonstration of direction and capability. The core ideas are real and running: permission-aware semantic search, RAG over Alfresco and Nuxeo, and on-premises LLM support.

What it is not yet is a finished open-source release. We are deliberately taking the time to ship something you can actually clone, run on your own infrastructure, and contribute to: with solid deployment, clear documentation, and proper licensing. In other words: code, not slides.

That work is in progress now, and the open-source release is targeted for later this year. Here’s the plan.

The roadmap

When What
June 2026 Technical preview published on this blog; live demo and Q&A at CommunityLIVE US.
Summer 2026 Finalizing the open-source release: deployment, quickstart, demo data, architecture docs, licensing.
September 2026 (target) Open-source release. The Content Lake stack (Alfresco + Nuxeo ingesters, hxpr integration, RAG service, UI extensions) published on GitHub. Clone it, run it on your own infrastructure, contribute.
September 2026 Virtual developer meetup & hands-on workshop, a launch session with a live walkthrough and Q&A. Open to the community; registration announced beforehand.
November 2026 Virtual developer meetup (APAC-friendly), panel, community Q&A, and a time-zone-friendly session for the Asia-Pacific region.

Dates are targets for a project still in active development; we’ll confirm specifics as we get closer. We would rather release when it’s genuinely ready than ship something half-finished to hit a date.

What you can build today: the on-premises AI story for Alfresco

You don’t have to wait for September to start. Content Lake is the most complete expression of an on-premises AI path for Alfresco, but it sits on top of building blocks that are already open source and available now:

And the Content Lake preview post itself walks through the architecture, the permission model, and the two-phase sync pipeline.

The bigger picture: start your AI journey on-premises with open source. Content Lake (September) adds a permission-aware, dual-repository semantic search and RAG layer with local LLM support: the self-managed entry point for customers constrained by privacy, data sovereignty, or infrastructure. As your needs grow (more scale, more accuracy, more advanced capabilities) there’s a natural path toward Hyland AI services and Content Intelligence Cloud (CIC).

Why this approach matters

This isn’t “AI bolted onto Alfresco.” The design choices behind Content Lake are deliberate, and each one answers a concern we hear repeatedly from the Alfresco community:

  • On-premises and self-managed. Run the entire stack (search, embeddings, and the LLM) on your own infrastructure, with local models via Docker Model Runner. No requirement to send content to an external AI service. For regulated industries and air-gapped environments, this is the difference between “can adopt AI” and “cannot.”
  • Data sovereignty by default. Your content never has to leave your network. This is built for the EMEA reality (GDPR, sector-specific compliance, and national data-residency rules) where “where does my data physically live?” is the first question, not the last.
  • Multi-source, not Alfresco-only. A single permission-aware AI layer spans both Alfresco and Nuxeo repositories. One semantic search and RAG experience across content that today lives in separate silos: no per-repository AI project, no duplicated pipelines.
  • Permission-aware: security is not an afterthought. ACLs are enforced server-side at query time. Users only ever retrieve, and get answers grounded in, content they’re already allowed to see. RAG that quietly leaks restricted documents is a non-starter; this avoids it by design.
  • Repositories stay the source of truth. Content Lake indexes and augments; it does not become a second copy of record. Minimal duplication, and your existing governance, retention, and lifecycle rules keep applying.
  • Open source: inspect it, run it, extend it, contribute. No black box. You can read the code, validate the security model yourself, adapt it to your environment, and shape its direction. That’s a real differentiator for organizations that won’t deploy what they can’t audit.
  • A growth path, not a dead end. Begin on-premises with open source today; evolve toward Hyland AI services and CIC when you need greater scale, accuracy, or advanced capabilities. You’re never forced to re-platform to move up.

In short: it lets existing Alfresco (and Nuxeo) customers begin their AI journey on their own terms (on their own infrastructure, under their own compliance regime, across their own repositories) without waiting for a cloud mandate or handing their content to a third party.

Want to engage before September?

Some of you need earlier visibility: for planning, proof-of-concept work, or a live customer opportunity. I’m glad to help directly.

I can arrange private demos and technical discussions for developers, partners, and customers who want to:

  • see a live walkthrough of the current state of Content Lake,
  • talk through how it fits an on-premises or hybrid architecture,
  • understand the roadmap and the expected feature set at release, or
  • explore early-access and pilot possibilities.

Reach out: I’ll schedule a call or send a recorded demo. If you’re an existing Alfresco customer with on-premises requirements, I’d genuinely like to hear what you need; that feedback shapes the release.

In short

  • Content Lake / hxpr was previewed at CommunityLIVE US and is in active development.
  • Open-source release targeted for September 2026, on GitHub: runnable and contributable.
  • Virtual meetups in September (launch + workshop) and November (APAC-friendly).
  • Today: Alfresco Simple RAG, the Transform Engines, MCP, and Quantum TLS give you real, open-source building blocks for on-premises AI.
  • Need it sooner? Contact me for a private demo or technical discussion.

We’re building this in the open, and we’re building it to last. Thanks for staying engaged, more soon.

Copyright © 2026 Khoros, LLC