Glean

Permissions-aware enterprise search and agent platform — the intelligence layer beneath workplace AI.

1. Core Product / Service

Glean is an enterprise "Work AI" platform built on a permissions-aware knowledge graph that indexes content across 100+ workplace tools (Microsoft 365, Google Workspace, Slack, Salesforce, Jira, etc.). Its original wedge was enterprise search: vector/semantic retrieval over company data using deep-learning LLMs, returning real-time, permissions-filtered results so each user sees only what they are authorized to access (glean.com/product/overview, 2026-06-29).

The platform has since expanded into three layers: enterprise search, Glean Assistant (a proactive coworker that surfaces tasks, drafts first passes, and answers grounded in personal and enterprise context), and Glean Agents (natural-language agents that plan, reason, and execute across the enterprise graph without predefined workflows). The May 2026 launch added Skills (codified repeatable workflows), multi-workstream delegation with approval controls, and Adaptive Reasoning for model/intelligence selection per task (glean.com/blog/may-2026-launch, 2026-06-29).

Glean emphasizes model neutrality — supporting 15+ LLMs to counter hyperscaler lock-in — plus governance SKUs (Glean Protect Plus), full observability of every query/answer/action, and an Agentic Engine plus a Canvas co-authoring UI (futurumgroup.com, 2026-06-29).

2. Target Users & Pain Points

Glean sells to large enterprises, including a meaningful share of the Fortune 500; enterprise deals can exceed $5M annually (futurumgroup.com, 2026-06-29). The core pain solved is knowledge fragmentation: employee data and answers are scattered across dozens of SaaS apps, each with its own permissions, so finding the right document or executing cross-app work is slow and error-prone.

Glean's differentiator is doing this while strictly respecting source-system permissions — a hard requirement in compliance-heavy environments where exposing the wrong document to the wrong employee is a security incident. As it moves into agents, the buyer expands from knowledge-management/IT leaders to platform teams who want to build and govern internal AI agents on top of a unified, permissioned enterprise graph.

3. Competitive Landscape

Glean's bet is to be the cross-stack, model-neutral intelligence layer beneath the interface, rather than win the chat UI itself — explicitly contrasting with Copilot, which is strongest inside the Microsoft ecosystem (techcrunch.com, 2026-06-29).

4. Unique Observations

• Glean's strategic shift is telling: from "better enterprise chatbot" to "connective tissue between models and enterprise systems." This is a deliberate retreat from the contested interface layer toward the harder-to-displace plumbing layer (permissioned knowledge graph + agent runtime). The wager is that the UI commoditizes while the permissioned retrieval substrate does not.
• The permissions-aware graph is the real moat — not the LLM. Any model can summarize; few systems can enforce per-user, per-source access controls across 100+ apps at retrieval time. This is exactly what hyperscaler bundles (Copilot, Gemini) struggle to replicate for non-native apps, and it is why model neutrality is a coherent strategy rather than a hedge.
• The Copilot distribution gap is existential, not cosmetic. With 90%+ of the Fortune 500 already on M365 and 20M+ paid Copilot seats, Glean must win on cross-stack breadth and governance depth before Microsoft closes the connector/permissions gap. The $5M+ deal sizes suggest Glean is succeeding at the top of the market, where heterogeneity (not pure Microsoft shops) is the norm.
• "250M+ agentic actions executed" (by Jan 2026) is the metric to watch — it reframes Glean from a search seat-license business to a consumption/agent-execution business, which is a fundamentally larger TAM than enterprise search ever was. Compare its agent ambitions to vertical agent peers like harvey (legal) — Glean is the horizontal version.

5. Financials / Funding

• Total raised (primary equity): $0.77B
• Latest valuation: $7.2B

6. People & Relationships

• Founders / key people: Arvind Jain (Founder & CEO; previously co-founded Rubrik and spent a decade at Google leading Search, Maps, and YouTube teams), Vishwanath T R, Tony Gentilcore, and Piyush Prahladka. Headquartered in Palo Alto, founded 2019 (glean.com/about, 2026-06-29).
• Notable investors: Kleiner Perkins, Lightspeed Venture Partners, General Catalyst, Sequoia Capital, Altimeter Capital, DST Global, Wellington Management (see §5).
• Partners / competitors: Integrates with 100+ enterprise tools (Microsoft 365, Google Workspace, Slack, Salesforce). Primary competitors are Microsoft 365 Copilot and Google Gemini (hyperscaler bundles), plus enterprise-search players Coveo and Guru; adjacent agentic/knowledge peers include hebbia and writer-ai.