spektr 3.0 Adds Real-Time Transaction Monitoring, Policy Controls and AI Casework

Compliance technology provider spektr launched spektr 3.0 on September 1, 2026, adding three products to its platform: Policy Engine, Transaction Monitoring and Case AI. The release extends spektr beyond onboarding and customer due diligence into ongoing monitoring and investigation workflows.

Transaction monitoring moves into the same customer record

The new Transaction Monitoring module lets compliance teams build detection rules for velocity, thresholds and multi-leg patterns through a visual rules interface. According to spektr, activity can be evaluated in real time, retrospectively or on a schedule. The system can block a transaction before completion or automatically open a case with the relevant transactions attached.

spektr says the monitoring engine is deterministic by design: each alert records the rule that fired, the values that breached it and the transactions that contributed. Because onboarding and transaction monitoring sit on the same platform, alert and case outcomes can update a customer’s risk profile, while onboarding attributes such as risk level, tenure and segment can feed into monitoring rules.

Policy Engine and Case AI

Policy Engine converts an institution’s own policy documents into proposed rules. The company says AI extracts individual requirements, but compliance staff must approve, edit or reject each rule before it takes effect. Approved rules are versioned and can be used to govern workflows, agents and case decisions.

Case AI is designed for sanctions, PEP and adverse-media reviews, alert triage and periodic reviews. Agents gather evidence, return a confidence level and recommended disposition, and cite the evidence used. spektr says evidence is drawn from an institution-approved allow-list of sources such as ComplyAdvantage, Kyckr and OpenCorporates, while free-form web scraping is disabled by default. Human sign-off remains required for final decisions.

The launch reflects a broader RegTech trend toward connecting onboarding, transaction monitoring and investigation in one controlled workflow. The key practical distinction in spektr’s design is that AI-assisted casework is paired with named rules, versioned policy controls and human approval rather than operating as an unconstrained decision layer.