The battle for enterprise AI is heating up. Microsoft is bundling Copilot into Workplace. Google is pushing Gemini into Workspace. OpenAI and Anthropic are promoting on to enterprises. Each SaaS vendor now ships an AI assistant.
Within the scramble for the interface, Glean is betting on one thing much less seen: changing into the intelligence layer beneath it.
Seven years in the past, Glean got down to be the Google for enterprise — an AI-powered search instrument designed to index and search throughout an organization’s SaaS instrument library, from Slack to Jira, Google Drive to Salesforce. At this time, the corporate’s technique has shifted from constructing a greater enterprise chatbot to changing into the connective tissue between fashions and enterprise techniques.
“The layer we constructed initially – a great search product – required us to deeply perceive individuals and the way they work and what their preferences are,” Jain advised TechCrunch on final week’s episode of Fairness, which we recorded at Net Summit Qatar. “All of that’s now changing into foundational when it comes to constructing top quality brokers.”
He says that whereas massive language fashions are highly effective, they’re additionally generic.
“The AI fashions themselves don’t actually perceive something about what you are promoting,” Jain stated. “They don’t know who the completely different individuals are, they don’t know what sort of work you do, what sort of merchandise you construct. So it’s a must to join the reasoning and generative energy of the fashions with the context inside your organization.”
Glean’s pitch is that it already maps that context and may sit between the mannequin and the enterprise knowledge.
The Glean Assistant is usually the entry level for purchasers — a well-known chat interface powered by a mixture of main proprietary (ie, ChatGPT, Gemini, Claude) and open-source fashions, grounded within the firm’s inside knowledge. However what retains clients, Jain argues, is all the pieces beneath it.
Techcrunch occasion
Boston, MA
|
June 23, 2026
First is mannequin entry. Slightly than forcing corporations to decide to a single LLM supplier, Glean acts because the abstraction layer, permitting enterprises to change between or mix fashions as capabilities evolve. That’s why Jain says he doesn’t see OpenAI, Anthropic, or Google as competitors, however reasonably as companions.
“Our product will get higher as a result of we’re in a position to leverage the innovation that they’re making available in the market,” Jain stated.
Second are the connectors. Glean integrates deeply with techniques like Slack, Jira, Salesforce, and Google Drive to map how info flows throughout them and allow brokers to behave inside these instruments.
And third, and maybe most necessary, is governance.
“You have to construct a permissions-aware governance layer and retrieval layer that is ready to carry the appropriate info, however understanding who’s asking that query in order that it filters the data primarily based on their entry rights,” Jain stated.
In massive organizations, that layer could be the distinction between piloting AI options and deploying them at scale. Enterprises can’t merely load all their inside knowledge right into a mannequin and create a wrapper to kind out the options later, says Jain.
Additionally essential is making certain the fashions don’t hallucinate. Jain says its system verifies mannequin outputs towards supply paperwork, generates line-by-line citations, and ensures that responses respect present entry rights.
The query is whether or not that center layer survives as platform giants push deeper into the stack. Microsoft and Google already management a lot of the enterprise workflow floor space, and so they’re hungry for extra. If Copilot or Gemini can entry the identical inside techniques with the identical permissions, does a standalone intelligence layer nonetheless matter?
Jain argues enterprises don’t wish to be locked right into a single mannequin or productiveness suite and would reasonably go for a impartial infrastructure layer reasonably than a vertically built-in assistant.
Buyers have purchased into that thesis. Glean raised a $150 million Collection F in June 2025, almost doubling its valuation to $7.2 billion. Not like the frontier AI labs, Glean doesn’t want large compute budgets.
“We now have a really wholesome, fast-growing enterprise,” Jain stated.
