Knowledge Assistant

An internal knowledge assistant helps employees ask questions and retrieve answers from company information, policies, and product knowledge.

Business impact

• Employee information search time — One deployment reported search time falling from over two minutes to 18 seconds.
• Call volume — One broker-facing pilot reported a 91% reduction in calls from those partners.
• Daily queries — One workplace assistant was described as handling 30,000 daily queries.

Data requirements

• Internal knowledge bases (text) — Used to supply answers and source-backed responses.
• Company documents (text) — Used for policies, product knowledge, and other internal information.
• Internal systems data (tabular) — One assistant searches across internal systems for employee queries.

AI methods and techniques

• Question answering — Employees ask questions in natural language and receive answers from internal knowledge sources.
• Conversational search — The assistant lets users retrieve information through chat instead of navigating separate systems.
• Generative AI — Several deployments use generative models to produce answers, summaries, or guided responses from company content.

AI models and model families

Amazon Bedrock, ChatGPT, Glean, HyperClova X, Claude Cowork, Copilot for Microsoft 365

Industries

Real-world evidence

25 documented case studies on record.

Companies using this: adidas, Bank of America, BlackRock, Coca-Cola, Deutsche Telekom, Fivetran, Home Depot, Honeywell, Hyundai, IBM, Infosys, Merck, NEC, Naver, Panasonic and 10 more.