Idea
System detecting and clustering real-time technical risk events from noisy customer incidents for enterprise cloud services.
Research Paper
Core Innovation
This paper presents TingIS, which combines efficient indexing with Large Language Models for multi-stage event linking to merge diverse user incident reports. It integrates cascaded routing for business attribution and a multi-dimensional noise reduction pipeline, enabling stable, real-time extraction of actionable incidents at enterprise scale.
Why It Matters
Large-scale cloud services face costly downtime and user trust loss from undetected technical anomalies. TingIS improves incident discovery speed and accuracy by filtering noise and linking diverse reports, enabling faster mitigation and reducing operational risks. Its scalable design supports high-throughput environments, transforming incident management workflows.
Market Size (TAM)
$10–20B TAM for enterprise IT incident management; $2–5B SAM from cloud providers and large enterprises. Driven by increasing cloud adoption and demand for real-time operational intelligence.
Potential Customers & Pain Points
- Cloud service providers – Need faster accurate anomaly detection
- Large enterprises – Struggle with noisy incident data
- IT operations teams – Require precise incident routing and prioritization.
Business Model
Subscription-based SaaS platform with tiered pricing based on message throughput and feature set, targeting cloud providers and large enterprises.
Competitive Landscape
- PagerDuty
- Splunk
- Datadog
- Moogsoft
Implementation Challenges
- Integration complexity with diverse enterprise systems
- Handling evolving incident patterns and noise
- Adoption resistance due to existing legacy tools
Validation Strategy
- Deploy pilot with select cloud service customers to measure alert latency and discovery rate improvements
- Benchmark against existing incident management tools on real-world noisy data
- Collect user feedback on routing accuracy and operational impact
Research Paper Overview
TingIS: Real-time Risk Event Discovery from Noisy Customer Incidents at Enterprise Scale
Summary
TingIS is an enterprise-grade system for real-time detection of technical anomalies from noisy customer incident data, achieving high accuracy and low latency at large scale. It extracts actionable incidents from diverse user reports using a multi-stage event linking engine with LLMs, noise reduction, and precise business routing, significantly improving incident discovery and response.