What

Variance, formerly Intrinsic, is an AI-based risk operations platform that uses autonomous agents to detect threats, investigate activity, and enforce mission-critical decisions. Based on the page, it is designed for teams handling financial fraud, marketplace compliance, threat intelligence, and trust and safety across content, transaction, and communication data.

The product appears positioned for enterprise and modern operations teams that need to connect fragmented data sources, apply policies in plain English, and automate investigations and enforcement. Its core workflow centers on ingesting data, mapping assets and structures, uploading policies or SOPs, detecting risks and coverage gaps at scale, and triggering automated mitigations through integrations or webhooks.

Features

Multi-source data connection: Variance connects disparate data streams, logs, assets, and behavioral signals to create a unified view of risk across systems.
Plain-English policy enforcement: Teams can bring their own policies and SOPs or use updated policy libraries, allowing AI to enforce operational rules without relying only on complex playbooks.
Autonomous investigations at scale: The platform is presented as handling both granular investigations and large-scale pattern detection across billions of transactions and petabytes of data.
Coverage gap detection: Variance identifies blind spots, new risk vectors, and gaps in detection coverage so teams can improve monitoring over time.
Workflow automation and mitigation: Through integrations, custom APIs, data lake syncs, and webhooks, AI agents can notify teams and take actions to mitigate risk.
External signal enrichment: The platform can enrich internal data with OSINT and other external signals to support proactive audits and investigations.

Helpful Tips

Validate data readiness early: Products like this depend heavily on clean access to logs, transaction data, communications, and asset definitions, so data mapping should be treated as a first implementation milestone.
Start with a narrow policy domain: Initial rollout is often smoother when teams begin with one high-priority use case such as fraud, abuse, or marketplace compliance before expanding policy coverage.
Review automation boundaries carefully: Because the platform is designed to enforce decisions and trigger mitigations, buyers should define which actions can run autonomously and which require human approval.
Assess policy maintenance workflows: Variance emphasizes plain-English SOPs and updated policy libraries, so decision-makers should confirm how internal teams will govern policy changes and exception handling.
Measure value through investigation speed and coverage: The most practical evaluation criteria for this category are usually reduced mitigation time, broader risk coverage, and faster case resolution rather than only detection volume.