Company knowledge as model context
Glean connects a company's documents, conversations and application records so employees can search across them and ask questions in natural language. Its AI role combines retrieval with language-model responses: the assistant works from connected workplace information and links answers back to relevant sources. The company's account says retrieval follows the user's existing permissions. [1]
Glean Chat's June 2023 announcement added document synthesis and content generation to that search foundation. A person could ask about a project's progress or summarize information spread across several applications. These are company-context tasks, where finding the appropriate internal evidence is part of producing a useful answer. [1]
From answers to configured actions
In June 2024, Glean made Apps, Actions and APIs generally available. Employees could configure specialized agents with natural-language instructions and selected knowledge sources; developers could build on search and chat APIs. Actions connected those agents to operations in other business software. [2]
The platform supported models from several providers, separating Glean's retrieval and enterprise context from the underlying language model. Its contribution is that connection between company information, model reasoning and permitted actions. The release describes the platform's controls; it does not establish that every deployed agent answers correctly or follows policy in every case. [2]