What we announced at Docebo Inspire 2026 and why it matters for your people.

At Docebo Inspire 2026, we announced the most significant release in our company’s history, stepping into a new category: a unified hub where learning, enterprise knowledge, and skills intelligence come together in one closed loop. In a category defined by legacy systems and decades of incremental change, we’re choosing to lead it forward.

Here’s what we announced.

Docebo AgentHub: From chatbots to agents that actually do things

Most AI in L&D today is reactive. You ask a question, you get an answer. That’s useful, but it’s not the same as having an intelligent system that knows your learners, knows your content, and can act on your behalf.

Docebo AgentHub changes that. These aren’t chatbots. They’re agents that reason, decide, and act, delivering personalized learning reminders, running skills campaigns, supporting content creation, and following through on learning programs without someone manually setting it all in motion. An admin can ask an agent to build a 15-minute onboarding micro-course for new sales reps, and the agent will pull from Drive, Confluence, SharePoint, and Docebo’s own catalog, generate the outline, the assessment, and the media plan, and deliver a ready-to-review draft. One-off knowledge becomes a course. A scattered wiki becomes a skills campaign. A product release note becomes an interactive podcast.

Docebo AgentHub was built on the foundation of our acquisition of Zive, a Germany-based enterprise AI and knowledge platform.

“Work doesn’t happen inside a single platform, and learning shouldn’t either,” said Piers Wermbter, Co-Founder and CTO, Zive. “With Docebo AgentHub and all the improvements to Docebo AI, we’re connecting Docebo to the full fabric of how enterprises operate so knowledge reaches people where they already are.”

We agree.

Enterprise Knowledge: All the knowledge your organization already has, finally usable

Knowledge inside an enterprise is scattered across Confluence, SharePoint, OneDrive, Google Drive, ticketing systems, CRM, and dozens of other tools. Learning has historically lived in a separate silo, disconnected from the knowledge people actually rely on to do their jobs.

Enterprise Knowledge closes that gap. It connects Docebo to 20+ knowledge sources, giving organizations a single system where their knowledge is accessible, searchable, and ready to act on. For the first time, one-off documents, policy pages, product specs, and tribal knowledge buried in shared drives become usable inputs for learning and workforce readiness.

Docebo Skills Intelligence: Closing the loop on workforce capability

Learning and knowledge only create value when an organization knows who needs what, when, and why. That’s the job of skills intelligence, and it’s the layer that makes everything else smarter.

Through the acquisition of 365Talents, Docebo now embeds a full skills intelligence engine directly into the platform. Skills campaigns surface roles, proficiencies, and inferred skills across the workforce. Gaps trigger targeted learning paths powered by Enterprise Knowledge and AgentHub. And completed learning feeds updated skill levels back into the system so the picture gets more accurate every day. Skills intelligence is no longer a separate system bolted on the side. It’s the connective tissue that makes every other Docebo capability work harder.

Docebo MCP: Your learning data, inside the AI tools your team already uses

Here’s a problem most organizations are quietly running into. They’ve rolled out AI assistants including Claude, Microsoft Copilot, and ChatGPT. People are using them every day. But those tools know nothing about your learning programs, your skills data, or your people.

The investment you’ve made in training? Invisible to the tools your workforce depends on most.

Docebo MCP (Model Context Protocol) Server fixes that. It makes Docebo a native knowledge source inside any MCP-enabled AI assistant. Your AI tools can now look up learner progress, surface course recommendations, and check certification status without a custom integration or months of development work.

Docebo MCP Server is live today in public beta, with full launch in July 2026.

Companion: Docebo, without switching tabs

Companion is a headless module that brings Docebo’s AI-powered learning directly into whatever your people are working on, without asking them to stop what they’re doing. This browser extension builds on the architecture of Docebo’s headless functionality to allow learning to happen anywhere learners work. No new window. No “I’ll look that up later.” Just learning, in context, in the flow of work.

Docebo AI Tutor: On-demand learning support, around the clock

Some people want to be guided through material. Some have a specific concept they need explained before a meeting. Some want a summary before they dive in. Docebo AI Tutor handles all of it. It’s an on-demand AI study companion that explains concepts, summarizes content, and guides learners through material whenever they need it, not just when an instructor or colleague happens to be available.

The platform improvements your admins have been asking for

Alongside the headline launches, we shipped a wave of improvements built directly from customer feedback. A few worth calling out: Interactive Podcasts turn lesson content into dynamic episodes with two AI hosts, to bring learning right into a format that’s already known and loved. Lesson Narrator adds natural AI voiceover to any lesson, so learners can consume content on the move, in their own language. Brand/Experience Studio gives organizations full control over platform branding including colors, fonts, and logos without touching a line of code.

We also rebuilt the admin navigation experience, improved eCommerce with better shared credit management and multi-currency support, overhauled the Content Marketplace, and made meaningful improvements to live training management, certification lifecycles, enrollment rules, and analytics.

Why this release is different

We move fast. In the last quarter alone, we increased new product releases by 50% or more and most of our code is now AI-assisted. That pace made this release possible.

But speed isn’t the point. The point is having a clear view of where enterprise learning needs to go and building the infrastructure before customers hit the wall.

Skills gaps are real. The tools people use every day are changing. And L&D teams are being asked to do more with the same resources. The only way that math works is if the platform does more of the heavy lifting.

That’s what we built. A learning platform that works in the flow of work, not alongside it.

Want to see what’s new up close? Request a demo or explore the release notes.