AIOps for banking platforms: Telemetry to triage

Modern banking platforms generate huge volumes of logs, metrics, traces, and events, yet incident response still stalls when signals are noisy, incomplete, and disconnected from service context. This session walks through a practical AIOps approach that takes teams from raw telemetry to production-grade triage. The speaker covers how to standardize telemetry pipelines, define "minimum viable signals," and enrich data with service topology, ownership, and change history so anomalies are actionable, not just detected.

You’ll learn patterns to reduce alert fatigue, how to keep the solution auditable for regulated environments, and implementation tips for operating ML as a platform capability.