Miami – October 01, 2026 -- Nuix used Sibos 2026 to position its Nuix Neo platform as a tool for financial institutions to convert unstructured data into usable financial crime intelligence, citing that up to 90% of enterprise data remains unreadable by conventional systems.
Nuix Neo processes over 1,000 file types to surface hidden risk signals
The platform's patented processing engine extracts information from emails, chat logs, counterparty data, transaction records, case files and voice recordings at a binary level. Deterministic natural language processing classifies and risk-scores the data, while semantic search surfaces relevant documents across languages based on meaning rather than keywords alone.
"AI is only as good as the data beneath it, and in complex financial crime investigations seeing and connecting all evidence is critical," said Chris Stephenson, Head of Industry Advocacy at Nuix.
Linkurious acquisition adds graph analytics to map transaction networks
Nuix has integrated decision intelligence and graph analytics capabilities from its Linkurious acquisition directly into Nuix Neo. The addition lets financial crime teams visualize relationships across accounts, people, organizations and transactions, enabling analysts to trace multi-hop connections and overlay external enrichment sources or corporate registries to expose criminal schemes.
Stephenson said the platform allows analysts to follow a counterparty across accounts, jurisdictions and years of transaction history and demonstrate to a regulator exactly how a conclusion was reached. He noted that financial crime teams are not short of alerts but short of certainty, as fragmented and unstructured data forces manual piecing-together of context.
Historical transaction data can be reprocessed to catch patterns earlier models missed
Nuix said compliance teams can reprocess years of historical transaction data through Neo to surface patterns that earlier models failed to detect, while triaging current alerts with greater confidence. The system records every processing action at the point it occurs rather than reconstructing it afterward, intended to support audit trails for regulators, Financial Intelligence Units or courts.
Platform supports BYO-AI while flagging outputs as probabilistic, not final
Nuix Neo's architecture uses deterministic scoring aimed at explainable and auditable outcomes, and organizations can connect their own approved large language models via a BYO-AI capability to maintain data sovereignty. Nuix cautioned that generative AI outputs remain probabilistic and must be verified against source material, intended to support rather than replace the judgment of qualified investigators and legal professionals.