Enterprise Search

Employees search internal knowledge bases, documents, and work systems through natural-language query tools and chat assistants.

Business impact

• hours saved monthly — Confluent reported more than 15,000 hours saved monthly across employees using Glean.
• investigation time per ticket — Confluent reported Technical Support investigation time per ticket fell by 5–10 minutes.
• employee favorability rating — Autodesk reported an employee favorability rating of 86% for the search capability.
• questions asked per day — Cerebras reported 15,000 questions a day within three months of launch.
• time spent on knowledge retrieval tasks — Lenovo reported a 30% reduction in time spent on knowledge retrieval tasks.
• employees reporting reduced time spent searching for information — Lenovo reported 81% of employees said they spent less time searching for information.
• AI agents created — Zillow reported more than 4,600 AI agents created using Glean.
• user adoption across departments — Stanley Black & Decker reported 8,000 users across departments.
• manual processing time — Stanley Black & Decker reported a 60% reduction in manual processing time for target workflows.

Data requirements

• internal documents (text) — Used as a source for searching and summarizing company knowledge.
• knowledge bases and wikis (text) — Connected repositories that store internal information.
• intranet content (text) — Searched for contextual workplace answers.
• project repositories and code-related knowledge (text) — Used in engineering and development knowledge search cases.
• technical support tickets (text) — Used for support investigation work.

AI methods and techniques

• retrieval-augmented generation — Used to pull relevant internal content into answers for employee queries.
• semantic search — Used to match employee questions with relevant internal knowledge beyond exact keyword search.
• chat-based question answering — Used to let employees ask for information in natural language.
• citation-supported retrieval — Used in one case to return cited answers from connected company tools.

AI models and model families

GPT for Business, OpenAI, LLMs

Industries

Real-world evidence

18 documented case studies on record.

Companies using this: ABB, Airbnb, Autodesk, Cerebras, Confluent, Daifuku, Mercedes-Benz, Honda, HubSpot, Lenovo, Mastercard, Mitsubishi Electric, Siemens Energy, Spotify, Stanley Black & Decker and 3 more.