Azure Capacity Forecaster
Forecasts weeks-to-quota-exhaustion from a synthetic subscription's usage trend using a transparent linear-regression model, extending the Cloud Capacity project.
Deterministic forecast + AI narrative — synthetic usage data, simple linear trend
The forecast is a simple linear-regression trend fit over the most recent six weeks of usage for the profile you select — not a sophisticated ML model, and the tool never claims otherwise. Weeks-to-exhaustion and the recommended new quota are computed deterministically; only the narrative urgency read and recommended action underneath are AI-generated. The four usage profiles are synthetic, illustrative examples of common capacity-growth patterns, not a real subscription's telemetry.
How it works
Pick one of four synthetic Azure usage profiles to see its current quota, latest utilization, weekly growth rate, and a linear-trend forecast of when it will exhaust the quota (fit on the most recent six weeks, since older history can mask a plateau) — then ask the capacity advisor agent how urgent a quota-increase request is.
Try it
Select a synthetic usage profile to forecast.
Loading profiles…
Pick a profile above to begin.
Why this exists
This extends the Cloud Capacity project record's theme of matching infrastructure decisions to real demand: the same question the quota platform answers at global scale — is there enough capacity for what's coming — applied here to one subscription's trend, with a transparent, inspectable forecast rather than a black box.
Production checklist
- Forecast is ordinary least-squares linear regression over recent weeks — no hidden ML model.
- Trend classification is quota-relative (0.3% of quota/week), so small quotas aren't mislabeled.
- The AI layer only narrates the already-computed forecast; it cannot alter any number.
- Dedicated Foundry prompt agent, versioned and isolated from the portfolio chatbot.
- Automated backend tests cover a near-limit profile and a plateaued profile.