Glean Feature Trails

Glean's agent orchestration platform has evolved to incorporate sandboxes, a wider range of actions, enhanced personalization, and a sophisticated context engineering layer. This layer unifies and continuously learns context from purpose-built systems like search, graph-based structures, and enterprise memory. It leverages specialized indexes for different data types (calendar, documents, structured data), employs hybrid connectors with data modeling for heterogeneous enterprise data, and utilizes agent identity for autonomous operation with scoped service credentials, enabling agents to act through their own service accounts in various platforms like Slack, Jira, and GitHub. This improves security, auditability, and reliability of automated tasks.

Glean's enterprise AI stack continues to evolve, with a focus on integrating its knowledge graph and vector database capabilities with external platforms via the Model Context Protocol (MCP). This integration allows external tools like v0 by Vercel to access Glean's permission-aware, grounded context, enabling AI-generated applications to be informed by live company knowledge, including documentation, systems, and decisions. The system ensures consistent, permission-aware answers by respecting existing data permissions. This post details the integration of Databricks as a native data source within Glean Assistant, enabling users to query Databricks data using natural language or read-only SQL. Glean translates natural language queries into SQL via Databricks Genie and combines these insights with broader enterprise context from documents, tickets, and other sources. The system also supports building polished artifacts like charts and slides from the analyzed data. Further integrations include indexing Databricks AI/BI dashboards and Genie Spaces, direct SQL querying within Glean Assistant and Agents, and bringing Glean's enterprise context into Databricks workflows via the MCP.

This post details the integration of Glean meeting notes into the existing unstructured data governance framework. It enhances the framework by transcribing conversations, allowing note-taking alongside transcripts, and generating post-meeting summaries focused on key decisions and action items. These meeting notes are made searchable and referenceable alongside other enterprise context (docs, messages, tickets, CRM records), becoming first-class artifacts within Glean. The system ensures that meeting transcript connectors for sales teams (Zoom, Granola, Fathom, Otter) and sales intelligence connectors (Salesloft Conversation Intelligence) and sales enablement connectors (Crayon, Showpad) are integrated, expanding Glean's native coverage across sales tools and making more conversation context searchable and usable in Assistant, answers, and workflows. This extends connected context across even more of the systems sales teams use every day.

Glean's security AI agents framework has evolved to address the challenges of traditional security architectures by providing a structured approach to evaluating and governing AI agents. This includes the AWARE framework, which offers shared decision criteria for CIOs and CISOs, focusing on intent, context, guardrails, runtime risk, and observability. The framework enables a repeatable process for assessing agent risk, moving from 'risk is too high for now' to approval by establishing clear technical and operational guardrails. This post details seven essential guardrail decisions for deploying enterprise AI agents successfully, covering model choice and hosting, decision architecture and retrieval-augmented generation, tool permissions and role-based access control, human-in-the-loop and risk tiers, observability and audit logging, input and output validation with content safety, and cost controls and financial operations.

This post introduces and explains the concept of webhooks, detailing their technical components, how they differ from API polling, and their benefits. It covers the event source, event, webhook endpoint, event payload, subscriber, HTTP request/response, security, and delivery/retry mechanisms. Use cases such as monitoring, incident management, infrastructure automation, event-driven automation, real-time notifications, and data integration are also discussed.

Glean's design system has been revamped to prioritize accessibility by default. This includes standardizing components with opinionated APIs and design tokens, ensuring consistent behavior and appearance across the product. The system now leverages Base UI for out-of-the-box screen reader and keyboard accessibility, with components exposing correct roles, states, and relationships. Focus management for dialogs and interactive elements has been improved, reducing the likelihood of accessibility regressions. Glean has also completed its VPAT, providing customers with a detailed report on its compliance with WCAG standards.

Glean has introduced code search and code writer capabilities to enhance engineering workflows. Code search provides deep context from all code repositories, enabling engineers to understand code, explain implementations, generate architecture diagrams, and update documentation. Code writer facilitates the transition from understanding to implementation by generating pull requests that adhere to existing patterns, reusing code and tests, and running checks. These features are integrated into existing workflows. This update enhances the Assistant with Search in Chat, autocomplete, and image generation, creating a unified workspace for discovery and execution grounded in enterprise context. Search in Chat and autocomplete integrate search results and document suggestions directly into chat, powered by trusted, permissions-aware search. Image generation allows for the creation of visuals directly within chat, grounded in enterprise documents, tickets, code, and conversations, with model-agnostic support and options for downloading generated images.