Glen

Glen is a shared memory layer for organizations, designed to capture and store knowledge from both human work and AI agent interactions. As agents and people perform tasks, Glen automatically records observations—decisions, processes, preferences, and facts—into a single, persistent store. This enables every agent (coding assistants, support bots, cron jobs) to recall the same knowledge, eliminating silos and ensuring consistency across the organization. For example, once a support lead resolves a refund policy question, all agents can answer similarly, and a new hire's agent inherits the expertise of the most senior person. The product targets teams running multiple AI agents in production, especially those using MCP-compatible clients like Claude Code, Cursor, and Codex. Glen is model-agnostic and integrates via the Model Context Protocol (MCP), working with any agent that supports MCP. It is also designed for humans—anyone can query Glen through natural language prompts to surface past decisions, playbooks, or customer commitments. Key features include automatic observation capture, provenance tracking on every piece of knowledge, observation-level RBAC (access control per observation), and private mode for sensitive conversations. Glen offers a free tier with a $10/mo usage credit and paid plans with increased credits and features like SAML SSO, audit logs, and dedicated support. What sets Glen apart is its focus on organizational memory rather than individual session memory—knowledge persists beyond sessions and context windows, making it a shared asset that compounds over time. Unlike episodic memory products like Mem0 or single-tenant solutions, Glen treats all agents as one learning system.