UST BaselineIQ - Network KPI Dynamic Threshold Service

Traditional static thresholds quickly become inaccurate as traffic patterns, seasons, and network configurations change - generating noisy false alarms or missing genuine anomalies. UST BaselineIQ builds a dynamic, self-learning baseline for every KPI at every network element, reflecting true time-of-day, day-of-week, and seasonal behavior. As the network evolves, BaselineIQ's baselines evolve with it - automatically recalibrating without manual threshold tuning so anomaly detection stays accurate and alarm volumes stay meaningfully low.

BaselineIQ is designed for NOC managers, network engineers, and SRE teams at telecommunications operators and large enterprise network environments who need to reduce alarm noise and surface genuine performance anomalies across thousands of network elements.

• Per-node, per-KPI learning: Automatically builds individualized baselines from historical behavior for every monitored element
• Continuous adaptation: Baselines recalibrate as traffic patterns, seasonality, and network topology shift - no manual threshold tuning required
• Intelligent anomaly detection: Identifies genuine anomalies by comparing real-time KPI values against learned behavioral envelopes, cutting false-positive alarms
• Scalable architecture: Designed to handle large-scale telecom and enterprise networks with thousands of elements and millions of KPI streams

A mobile operator experiences seasonal traffic spikes during major events that trigger hundreds of false RAN alarms under static thresholds. With BaselineIQ, the system learns that elevated traffic during these periods is normal behavior, suppresses false alarms, and surfaces only genuine degradation events - enabling the NOC team to focus on real issues rather than chasing noise.

BaselineIQ is delivered as a professional services engagement structured in phases:

1. Discovery and Scoping: UST works with your team to identify target network elements, KPI sources, and integration requirements
2. Data Onboarding: Historical KPI data is ingested and validated to establish initial learning periods
3. Baseline Calibration: BaselineIQ trains per-element baselines and tunes sensitivity parameters
4. Go-Live and Validation: Baselines are activated in production with side-by-side comparison against existing thresholds
5. Handoff and Knowledge Transfer: Configured system, operational documentation, and trained staff

BaselineIQ can be deployed on AWS infrastructure leveraging services such as Amazon S3 for KPI data storage, Amazon SageMaker for model training and inference, and Amazon CloudWatch for operational monitoring of the BaselineIQ platform itself.

• Minimum historical KPI data: typically 4-8 weeks of per-element time-series data
• Network types: mobile RAN, transport, core, and fixed-line access networks
• Buyer responsibilities: provide network element inventory, KPI data access, and stakeholder availability during discovery
• Out of scope: hardware provisioning, network design changes, and third-party NMS licensing

UST is a global digital transformation company delivering technology services and solutions across industries including telecommunications. BaselineIQ leverages UST's deep expertise in network operations, machine learning, and telecom domain knowledge.

To get started, subscribe through AWS Marketplace and a UST engagement manager will reach out to schedule your discovery session.