As competition intensifies in the enterprise artificial intelligence market, major technology companies are racing to control the user interface. Microsoft is bundling Copilot into Office, Google is embedding Gemini across Workspace, and AI developers such as OpenAI and Anthropic are selling directly to corporate customers. Nearly every software-as-a-service provider now offers an AI assistant.
Glean is taking a different approach.
Rather than competing for the interface, the company is positioning itself as the intelligence layer that connects AI models with enterprise systems and data.
Founded seven years ago, Glean initially set out to build an AI-powered enterprise search engine, indexing information across workplace tools such as Slack, Jira, Google Drive and Salesforce. Today, its strategy has evolved toward serving as the connective infrastructure that allows AI agents to operate effectively inside organizations.
Arvind Jain, Glean’s founder and chief executive, said the company’s early focus on search required a deep understanding of how people work and how information flows inside businesses.
“That foundation is now critical to building high-quality AI agents,” Jain said in an interview with TechCrunch recorded at Web Summit Qatar.
Jain said large language models are powerful but lack the context required to function effectively in enterprise environments.
“The AI models themselves do not really understand your business,” he said. “They do not know who the people are, what kind of work you do or how your systems operate. You have to connect the reasoning power of the models with the context inside your company.”
Glean’s platform aims to provide that connection. While many customers first encounter the company through the Glean Assistant, a chat-based interface powered by a mix of proprietary and open-source models, Jain said the underlying infrastructure is what drives long-term adoption.
One pillar of that infrastructure is model flexibility. Glean acts as an abstraction layer, allowing enterprises to switch between or combine models from different providers as capabilities change. Jain said the company views AI labs such as OpenAI, Anthropic and Google as partners rather than competitors.
“Our product improves as they innovate,” he said.
Another pillar is deep integration with enterprise software. Glean connects directly to tools such as Slack, Jira, Salesforce and Google Drive, enabling AI agents to understand how information moves across systems and to take action within them.
Governance, however, is the most critical component, Jain said. Glean’s system is designed to be permissions-aware, ensuring that users only receive information they are authorized to access.
“You need a governance layer that understands who is asking the question and filters information accordingly,” he said.
That capability is often what determines whether AI tools remain experimental or can be deployed at scale, Jain added. He said Glean also focuses on reducing hallucinations by verifying model outputs against source documents, generating citations and enforcing existing access controls.
The company’s approach faces challenges as platform giants move deeper into the enterprise AI stack. Microsoft and Google already control much of the productivity software used by large organizations, raising questions about whether a standalone intelligence layer can endure.
Jain argues that enterprises are reluctant to lock themselves into a single model or productivity ecosystem and prefer a neutral layer that can operate across vendors.
Investors appear to agree. Glean raised $150 million in a Series F funding round in June 2025, nearly doubling its valuation to $7.2 billion. Unlike frontier AI model developers, the company does not require massive computing resources.
“We have a very healthy, fast-growing business,” Jain said.
