Resource Allocation

AI recommends and executes how cloud, network, beds, and staff are allocated so teams can respond to demand and bottlenecks faster.

Business impact

• Cloud spend savings — Komatsu reports cumulative public-cloud savings of more than USD 650,000 since implementation.
• Wait time — The Johns Hopkins article says bottlenecks are quickly identified to reduce wait time.
• Network performance — Verizon says the system is used to improve network performance by detecting bottlenecks and anomalies and adjusting allocation.

Data requirements

• Cloud infrastructure telemetry (tabular) — Used by Komatsu to monitor servers, storage, and databases.
• Admission and patient flow data (tabular) — Used by Johns Hopkins and AP-HP to forecast admissions and staffing needs.
• Network interface statistics (tabular) — Used by Verizon to monitor routers and network interfaces.
• Environmental and weather sensor data (sensor) — Used by Verizon as part of network monitoring.
• CPU usage data (tabular) — Used by Verizon to track router and infrastructure load.
• Billing records (tabular) — Used by Verizon as operational data for anomaly and trend detection.

AI methods and techniques

• Predictive analytics — Used to forecast demand, trends, or admission rates so capacity can be allocated ahead of time.
• Machine learning — Used to analyse operational data and generate predictions, alerts, or recommendations.
• Anomaly detection — Used in Verizon to spot network outliers and potential bottlenecks or outages.
• Automation — Used in Komatsu to carry out Azure resource actions based on AI recommendations.

Industries

Real-world evidence

5 documented case studies on record.

Companies using this: Komatsu, Johns Hopkins Hospital, Verizon, Assistance Publique-Hôpitaux de Paris (AP-HP), Snap.