[DigitalToday reporter Chi-gyu Hwang] A battle is under way for the enterprise software AI market.

Microsoft has embedded Copilot in Office, and Google has integrated Gemini into Workspace. Software-as-a-service (SaaS) companies with well-known names are also rolling out AI assistants as default features. Against that backdrop, enterprise search specialist Glean Technologies is drawing attention for focusing on building an “intelligence layer” that links AI models and corporate data, instead of a user-facing interface.

According to a recent TechCrunch report, Glean’s strategy starts from a simple message. LLMs are powerful but do not know what is inside a company. LLMs cannot grasp who is in the company, what products it makes, or how it works. Glean’s strategy is to fill that gap.

To do that, Glean has positioned model neutrality as a differentiator. It is focusing on serving as an abstraction layer that mixes major proprietary models and open-source models, including ChatGPT, Gemini and Claude. Companies are not tied to a specific LLM and can switch flexibly on the platform even as model performance changes. For that reason, Glean CEO Arvind Jain (아르빈드 제인) defines OpenAI, Anthropic and Google as partners rather than competitors. He believes their innovation makes Glean’s product stronger.

Permission-based governance is also a key term Glean emphasizes.

Glean identifies who is asking a question and filters results to provide only information that matches that user’s access rights. It verifies model responses against original documents and cites sources line by line to control hallucinations. The company emphasizes that such a layer is essential for large organizations to move AI from pilot stages to actual deployment.

Connectors are also cited as Glean’s strength. Through connectors, Glean integrates deeply with major internal systems such as Slack, Jira, Salesforce and Google Drive to understand information flows and support agents in taking action within the actual tools.

There is also a view that the need for a separate intelligence layer like Glean could weaken, as big tech companies Microsoft and Google are also digging deep into the stack between AI models and corporate data.

Glean’s answer is that enterprise choice matters. It explains that companies that do not want to be locked into a single model or a productivity suite will choose a neutral infrastructure layer. It says demand exists for a flexible structure rather than vertically integrated assistants. Investors also appear to have agreed with that vision so far. Glean raised $150 million in a Series F funding round in June 2025, lifting its company value to $7.2 billion.

Glean recently also updated Glean Assistant. Until now, Glean Assistant could be used only through a chat interface, but the update added real-time voice support, allowing users to use the assistant conversationally without typing.

Content creation features were also strengthened. Glean Assistant can automatically create documents and images using fonts, colors and logos that match a company’s brand. A collaboration tool called Canvas was also newly added. On automation, it now supports more than 100 actions by integrating with major platforms including Salesforce, Google Calendar, Asana, Canva, Jira, Confluence and GitHub.

Agent Sandbox is an isolated environment for security-sensitive work. It includes a command-line interface, a code interpreter and a file system to analyze large-scale data without context window limits, while also blocking the risk of data leakage. Agent templates allow users to preset repetitive tasks such as summarizing action items, scheduling plans and weekly performance summaries, so they can be delegated to the assistant.

Glean’s strategy also serves as a test bed for a hypothesis in the AI industry that context and governance may hold higher value than models. Powerful models are available to anyone, but building a layer that safely connects internal enterprise data, controls permissions and verifies results takes time and trust. If Glean leads the early race with such a layer, big tech’s push south with bundling may not go as planned. Glean’s next moves are drawing attention.