-
Why do results improve after we clean up source data?
This comes down to a simple principle: Glean can only reason over the signals your environment gives it. When the underlying content is fresher, less duplicated, and better structured, Search and Assistant have a clearer picture of what is authoritative, what is outdated, and what should rank highest. A lot of people think…
-
Glean Model Council
One of Glean's advantages is that it is model agnostic. It would be cool if there was a "Model Council" feature (similar to Perplexity) where Glean would concurrently send queries to different models and compare the outputs to decide which models' output is the highest quality.
-
Canvas Support in Glean Agent Workflow!
We've been building a multi-step RFP Proposal Builder agent and ran into a UX challenge. The agent walks users through a structured, feedback-driven workflow, extracting RFP objectives, categorizing them, designing an outline, and mapping it to standard content blocks, before producing a full proposal document. Each step…
-
Disclaimer for AI Images
It would be great to be able to add a disclaimer to images if generated by the assistant. I'd like to have some kind of disclaimer on AI generated images like, "this image is LLM generated, not approved by legal, marketing, or use externally."
-
How do successful customers phase their Glean rollout so it doesn’t overwhelm IT or end users?
The strongest rollouts usually do not start with a company-wide big bang. They start with a narrow first phase: a small set of high-value connectors, a clearly defined user group, and a few concrete use cases that can show value quickly. From there, successful teams expand in waves — adding more sources, more departments,…
-
Feature request: Native Microsoft Fabric MCP integration in Glean
Hi Glean team, We’re actively working on integrating Microsoft Fabric Data Agents via MCP with Glean and wanted to submit a feature request for a more first‑class experience. Right now, we can reach the Fabric MCP endpoint directly and confirm that: The Fabric MCP server is reachable and responds correctly via direct…
-
Agent Step: "Wait for user's response"
With Glean Agents, it would be helpful to have the ability to use a "wait for user's response" step, where the agent would pause until the user provides a response, then proceed to the next step in the sequence? This would mimic a chatbot, conversational type of experience. Example: We want to build an agent that will…
-
Merge or Unify Step to Combine Branches in Agent Steps
It would be very useful to have the ability to merge or unify branches in an agent's workflow. For example, if an agent needs to branch into different topics to handle a different user experience, actions or knowledge sources, but then at the end of the agent sequence, there are final "Respond" and "Create Google Doc"…
-
Allow mails to the registered domain for Glean
Users are trying to build workflows that inform users of updates over email which may be derived from the workflow Today, the Glean send email action only supports sending mails to the user running the agent, meaning this will not work Limitations Documented here: https://docs.glean.com/agents/actions/glean/email Are we…
-
Hubspot - Playbooks Object
I understand that Hubspot as a data source has limitations in indexing certain objects. However, one feature we would like to enable is the Playbooks object, which aligns with our operational goal of improving quality. To give some context, we plan to develop a Glean agent that reviews support tickets in Hubspot while…
