Knowledge graph - Scientific news
On October 22, 2026, investigative analytics and intelligence software provider Nuix (ASX: NXL) launched the Nuix Neo AI Agent, an agentic intelligence solution combining Large Language Models (LLMs) with enterprise knowledge graphs to automate high--stakes corporate investigations, regulatory reviews, and compliance audits.
Strategic Context & Business Value
Traditional LLM deployments struggle in regulated, high--liability enterprise contexts due to hallucination risks, non--deterministic reasoning, and an inability to navigate complex relational data across fragmented communication channels (e.g., chats, emails, and transaction logs). By grounding LLMs in Nuix's forensic--grade data layer and native knowledge graph, the platform addresses key adoption barriers:
- Automated Graph Reasoning: The agent automatically generates graph queries from plain--language user prompts, eliminating the need for technical query engineering while traversing complex multi--hop connections among individuals, accounts, communications, and transactions.
- Auditability & Evidentiary Provenance: Unlike black--box generative models, every insight is deterministically linked back to underlying source documents, generating a transparent audit trail of every intermediate analytical step suitable for court proceedings and regulatory defense.
- Operational Governance: Built on the Nuix Neo Gen AI Framework, the system enforces granular role--based access control (RBAC), prompt guardrails, and centralized API spend controls to mitigate exposure to enterprise risk and data leakage.