Knowledge Base Management

Organises fragmented internal knowledge into connected, searchable context so employees can find answers and reuse expertise.

Business impact

• Model creation speed — Maana says its Knowledge Assistants help customers create models 3–10 times faster than when using other technology.
• Maintenance costs — Shell says the corrosion-related application aims to reduce maintenance costs.
• Unplanned downtime — Shell says the corrosion-related application aims to reduce unplanned downtime.

Data requirements

• Internal documents (text) — Technical papers, reports, emails, and other unstructured company documents are analysed.
• Relational databases (tabular) — DZDconnect combines graph databases with relational data.
• Internal company data (tabular) — Shell compiles and analyses company data from different parts of the business.
• Employee expertise — Maana’s knowledge graph captures people’s expertise and understanding of business processes.

AI methods and techniques

• Knowledge graphs — Used to represent and join data, context, and expertise across silos.
• Natural language processing — Used by Shell to extract information from unstructured documents and support knowledge graphs.
• Information extraction — DocAssist extracts relevant content from PDFs, technical papers, reports, and emails.
• GraphRAG — Neo4j describes it as a way to retrieve context from connected data for trustworthy AI.
• Agentic AI — Merck says Neo4j supports its agentic AI strategy for R&D knowledge work.

Industries

Real-world evidence

4 documented case studies on record.

Companies using this: Fujitsu, Shell, Deutsches Zentrum für Diabetesforschung (DZD), Merck Group.