The search engine for your own company - and, increasingly, the place where its AI agents come to work.
The mark hides a search icon and a smile. Behind it: a live map of every document, message, and permission in a company - and the machinery to reason across all of it.
Here is a problem that sounds too boring to be worth $7.2 billion: people cannot find things at work. The quarterly plan is in someone's Google Doc. The customer's contract is in Salesforce. The decision that reversed the plan is in a Slack thread from March. Nobody can see all three at once, and the AI everyone bought last year mostly can't either.
Glean's founders noticed this the way you notice a leaky roof - not all at once, but in the accumulated annoyance of watching it happen every day. Arvind Jain spent over a decade at Google as a distinguished engineer, working on Search, Maps, and YouTube, then co-founded Rubrik, a cloud data-security company that went public in 2024. As Rubrik scaled, Jain kept watching his own employees lose hours hunting for information they knew existed somewhere. He had spent years making the world's information findable. His own company's information was a mess.
So in 2019 he and a small group of former Google engineers - including Tony Gentilcore, who had worked on Google's instant-search technology, and T.R. Vishwanath - started Glean. The pitch was almost aggressively unglamorous: point Google-quality search at a company's own scattered apps, respect the permissions that already exist so nobody sees what they shouldn't, and make the whole thing feel like one place. This is the sort of idea that is easy to describe and genuinely hard to build, which is usually where the good businesses hide.
"Glean Assistant gives every employee a powerful enterprise AI assistant that connects to and understands company data and internet data."
The timing turned out to be excellent, though not for the reason the founders originally planned. Glean spent its early years building the unsexy plumbing - over 100 connectors to tools like Google Workspace, Microsoft 365, Slack, Jira, Confluence, Salesforce, and GitHub, plus a permissions model that decides, per user, what each search is allowed to return. Then large language models arrived, and it turned out that all of that plumbing was exactly what a corporate AI assistant needed underneath it. A chatbot without your company's context is a party trick. A chatbot wired into everything, with permissions intact, is something an enterprise will pay real money for.
And they did. Glean crossed $100 million in annual recurring revenue in the fiscal year ending January 31, 2025 - less than three years after launching the product commercially. Then it did the thing that makes investors lose their composure: it reached roughly $200 million ARR by December 2025, doubling in about nine months. Revenue that grows like that tends to attract capital, and capital duly arrived.
In September 2024, investors valued Glean at $4.6 billion. In June 2025, they valued it at $7.2 billion. The intervening event was not a new product so much as a new level of proof that enterprises would keep paying.
| Round | Amount | Date | Notable investors |
|---|---|---|---|
| Early rounds | ~$155M | 2019–2022 | Kleiner Perkins, Lightspeed, General Catalyst, Sequoia |
| Series D | $100M | Feb 2024 | Kleiner Perkins, Lightspeed, Sequoia |
| Series E | $260M | Sep 2024 | Altimeter, DST Global, Coatue (at $4.6B) |
| Series F | $150M | Jun 2025 | Wellington (lead), Khosla, ICONIQ, IVP, Sequoia (at $7.2B) |
The Series F reads like a roll call of enterprise-software investing: Wellington Management led, joined by Khosla Ventures, Bicycle Capital, Geodesic Capital, and Archerman Capital, alongside returning backers Altimeter, Capital One Ventures, Citi, Coatue, DST Global, General Catalyst, ICONIQ, IVP, Kleiner Perkins, Latitude, Lightspeed, Sapphire, and Sequoia. When that many firms keep re-upping, they are not betting on the demo. They are betting that context - not the underlying model - is the scarce resource in enterprise AI.
Glean now calls itself a "Work AI platform." Stripped of the branding, it is a stack: a search layer, an assistant on top of it, agents on top of that, and a model of the company holding the whole thing up.
Permissions-aware search across 100+ connectors. Ask a question, get results drawn only from documents, messages, and data you're already allowed to see.
A company-wide AI assistant that reasons through a request, builds a multi-step plan, and coordinates sub-agents in parallel. The third generation personalizes answers per employee.
A no-code environment where any employee can build an AI agent by chatting. Give it a task; it uses context and reasoning to automate the work - no professional services required.
A continuously updated model of a company's people, projects, processes, and data. It's the reason Glean's AI knows how work actually happens, not just what words appear in a file.
In practice, that means an engineer can ask what broke a deploy and get the incident doc, the pull request, and the Slack post-mortem together. A salesperson can have an agent draft a call follow-up grounded in the actual account history. An HR lead can build an onboarding concierge without writing code. The unifying trick is boring and important: everything runs through permissions, so the AI is useful without becoming a data leak.
Glean sells to mid-market and enterprise companies, generally 500 employees and up. Initial contracts commonly run $100K–$500K a year; large Fortune 500 deals can exceed $5M annually.
The list skews toward companies that are themselves data-heavy and engineering-led - the sort of organizations most acutely aware that their own knowledge is scattered across fifty apps. Which is, not coincidentally, exactly the pain Glean's founders felt at Google and Rubrik in the first place.
In late 2025, Glean launched the Work AI Institute, a research collaborative with academics at Stanford, UC Berkeley, and Harvard. Its first Work AI Index, published in June 2026, contained a finding that a lesser company might have buried: 87% of digital workers use AI at work, 75% say it makes them more productive - and only 13% say it has significantly improved their organization's performance.
"Workers say AI saves 11 hours a week - over a quarter of the workweek - but a lack of context is eating the gains."
The report also introduced a wonderfully bleak term: "botsitting." Workers, it found, spend about 6.4 hours a week supervising AI - more time than they spend using it to actually produce work. This is either a devastating admission or a very clever piece of marketing, and it is probably both. Glean's entire argument is that the bottleneck in enterprise AI is not the model's intelligence but its ignorance of your company. A tool that supplies that context - and Glean would like you to know it sells one - is how you close the gap between 87% adoption and 13% impact.
Ex-Google engineers led by Arvind Jain start the company (initially "Scio") to build AI-powered enterprise search.
Permissions-aware search ships, connecting a company's SaaS tools and documents into one layer.
A generative-AI assistant lands on top of search, and Glean reframes itself as "Work AI."
A $260M round in September as ARR climbs and headcount nearly doubles inside a year.
Raises $150M at $7.2B, launches Glean Agents and the Enterprise Graph, and crosses $200M ARR.
Publishes the Work AI Index and is named a Market Shaper for no-code agent builders.
Ex-Google distinguished engineer (Search, Maps, YouTube); co-founded Rubrik. Started Glean after watching employees lose hours hunting for information.
Product engineering lead. Worked on Google's instant-search technology before helping start Glean.
Technical infrastructure lead, another Google veteran on the founding team.
Glean is an enterprise "Work AI" platform that connects to a company's apps and documents to power permissions-aware search, an AI assistant, and no-code AI agents that automate work.
Glean was founded in 2019 by former Google engineers led by Arvind Jain, alongside co-founders including Tony Gentilcore and T.R. Vishwanath.
Glean was valued at $7.2 billion in its June 2025 Series F, up from $4.6 billion in September 2024. It has raised more than $780 million in total.
Enterprises such as Databricks, Duolingo, Canva, Confluent, T-Mobile, Reddit, Instacart, and Grammarly - typically companies with 500 or more employees.
Glean is model-agnostic and built around a permissions-aware Enterprise Graph spanning 100+ connectors, aiming to give AI accurate company context rather than tying customers to a single vendor's model or app suite.