TingIS: Real-Time Risk Event Discovery from Noisy Customer Incidents at Enterprise Scale
A new system called TingIS enables real-time detection of technical anomalies from noisy customer incident data at enterprise scale. The system uses a multi-stage event linking engine that combines efficient indexing with Large Language Models (LLMs) to merge events and extract actionable incidents from diverse user descriptions. A cascaded routing mechanism ensures precise business attribution. The paper is published on arXiv under identifier 2604.21889.
Key facts
- TingIS is an end-to-end system for enterprise-grade incident discovery.
- It uses a multi-stage event linking engine with LLMs for event merging.
- The system extracts actionable incidents from a handful of diverse user descriptions.
- A cascaded routing mechanism provides precise business attribution.
- Real-time detection of technical anomalies is critical for large-scale cloud-native services.
- Minutes of downtime can cause massive financial losses and diminished user trust.
- Customer incidents are a vital signal for discovering risks missed by monitoring.
- The paper is available on arXiv with ID 2604.21889.
Entities
Institutions
- arXiv