INETCO launches BullzAI Investigate with on-premise agentic AI for fraud alert triage

INETCO has launched BullzAI Investigate, a new module for banks and payment service providers that uses agentic AI to triage payment fraud alerts and assign explainable risk scores. The point is less “AI for AI’s sake” and more a practical one: cut alert-handling time without pushing transaction data outside the customer’s infrastructure.

1. INETCO, a Vancouver-based payment fraud prevention software company, says BullzAI Investigate sits alongside its existing BullzAI monitoring platform and is available now. The module is aimed at fraud case management, which puts it squarely in the part of the stack where PSPs and banks try to separate noise from cases worth an analyst’s time.
2. The product uses a proprietary small language model deployed entirely on-premise, so transaction data does not leave the customer’s infrastructure. INETCO says specialised AI agents collate transaction datasets, identify behavioural patterns, and automatically triage incoming alerts, surfacing the highest-risk cases for review.
3. Each recommendation includes an explainable risk score, which INETCO says gives investigators auditable reasoning for every decision. The system also uses a human-in-the-loop supervised learning cycle, taking analyst feedback after each closed case and feeding that back into the model’s output.
4. INETCO says early results show investigation times dropping from 10 to 30 minutes to approximately 20 seconds, which it describes as a 97 to 99 per cent reduction. The company also cites a recommendation precision rate of approximately 95 per cent, although it has not independently published the methodology behind that figure.
5. Ugan Naidoo, chief technology officer at INETCO, said the system “automates the heavy lifting” by collating transaction data, triaging alerts and delivering explainable risk scores, while keeping human oversight in control. INETCO also points to a Chartis Research report featuring a BullzAI deployment at an unnamed South African bank as third-party evidence, though the release does not include the full Chartis methodology or the bank’s deployment scale.

For high-risk operators and their PSPs, the important detail is the deployment model: on-premise remains a selling point where cloud-based fraud tooling runs into data localisation rules or internal bank policy. INETCO’s pitch sits in the same lane as other AI-assisted case management layers from vendors such as NICE Actimize, SAS and Featurespace, but the differentiator here is explicit data residency rather than a promise to make fraud operations “smarter” in the abstract.