AWS uses its own captured charge days. Actuals through 2026-03-30 UTC.

Forecast

Precomputed Effective Usage cost in USD, with daily expected ranges.

AWS · actuals and 14-day forecast

Light shading shows the 95% range; darker shading shows 80%. The divider marks the last actual on 2026-03-30. Coral markers flag daily spikes and drops outside the 95% range.

Show data (44 rows)
AWS · actuals and 14-day forecast
DayActual costExpected costForecast costLower95 costUpper95 costLower80 costUpper80 costAnomaly
2026-03-01$12.04$0.00–$0.00$12.43$0.00$1.11–
2026-03-02$7.00$0.75–$0.00$38.62$0.00$25.92–
2026-03-03$15.20$17.73–$0.00$56.54$0.00$41.70–
2026-03-04$8.00$18.42–$0.00$56.24$0.00$44.07–
2026-03-05$10.00$19.99–$0.00$60.93$0.00$46.26–
2026-03-06$10.00$15.06–$0.00$51.11$0.00$39.34–
2026-03-07$0.00$0.00–$0.00$38.40$0.00$24.17–
2026-03-08$0.00$1.57–$0.00$37.84$0.00$26.66–
2026-03-09$26.00$27.40–$0.00$62.95$3.01$51.88–
2026-03-10$17.00$44.39–$5.93$83.44$19.40$70.76–
2026-03-11$22.88$45.07–$7.54$83.78$20.53$71.18–
2026-03-12$52.00$46.64–$8.99$82.92$22.38$71.35–
2026-03-13$41.00$41.71–$3.06$79.53$16.26$67.50–
2026-03-14$59.60$26.06–$0.00$64.21$0.00$51.95–
2026-03-15$36.00$28.22–$0.00$66.24$5.23$52.75–
2026-03-16$63.00$54.06–$14.85$90.27$27.93$77.67–
2026-03-17$42.00$71.04–$34.08$113.68$44.86$96.72–
2026-03-18$83.00$71.72–$32.42$113.55$46.23$96.77–
2026-03-19$63.00$73.29–$32.36$109.49$48.16$97.07–
2026-03-20$65.20$68.36–$30.06$105.81$41.75$91.95–
2026-03-21$39.20$52.71–$15.27$87.90$27.05$76.37–
2026-03-22$26.00$54.87–$15.49$93.22$29.30$79.80–
2026-03-23$68.00$80.71–$41.60$119.99$55.31$104.67–
2026-03-24$140.00$97.69–$60.59$135.55$73.51$123.47spike
2026-03-25$139.00$98.37–$61.66$134.80$73.92$124.46spike
2026-03-26$97.00$99.95–$61.68$137.65$74.53$123.21–
2026-03-27$117.00$95.01–$57.20$131.74$69.93$118.99–
2026-03-28$53.00$79.36–$40.58$117.32$52.57$103.64–
2026-03-29$64.00$81.53–$41.97$125.00$56.08$106.42–
2026-03-30$117.20$107.36$107.36$71.47$147.14$83.83$133.65–
2026-03-31–$124.34$124.34$82.95$164.34$98.78$148.68–
2026-04-01–$125.03$125.03$83.91$164.63$99.34$148.83–
2026-04-02–$126.60$126.60$90.27$167.15$100.15$152.00–
2026-04-03–$121.66$121.66$77.35$157.99$93.09$145.26–
2026-04-04–$106.01$106.01$67.62$144.06$81.77$130.02–
2026-04-05–$108.18$108.18$73.56$145.14$84.24$133.82–
2026-04-06–$134.01$134.01$94.46$174.56$109.27$159.15–
2026-04-07–$151.00$151.00$113.85$189.31$126.60$174.15–
2026-04-08–$151.68$151.68$112.71$190.76$127.46$177.09–
2026-04-09–$153.25$153.25$115.35$190.14$128.39$178.51–
2026-04-10–$148.32$148.32$112.07$183.40$124.29$173.26–
2026-04-11–$132.67$132.67$95.43$172.39$108.54$158.89–
2026-04-12–$134.83$134.83$97.26$172.62$110.89$159.59–
2026-04-13–$160.66$160.66$123.41$199.32$134.42$185.42–

44 rows. Scroll inside the table to see more.

Projected 14-day Usage cost
$1,878.242026-03-31 → 2026-04-13 UTC
95% projected total range
$1,727.48 – $2,027.62Range for the sum of forecast days
Flagged days
2Observed cost outside the 95% expected range
Holdout RMSE
$40.797 observed days in the final-week backtest

Anomalies · AWS

AWS daily anomaly flags (USD)
DayActualExpectedExpected low (95%)Expected high (95%)DeviationSeverityDirection
2026-03-24$140.00$97.69$60.59$135.55$42.310.06spike
2026-03-25$139.00$98.37$61.66$134.80$40.630.06spike

2 rows. Scroll inside the table to see more.

Model card · AWS

AWS Prophet model configuration
SettingValue
Training window (UTC)2026-03-01 → 2026-03-30
Observed / unknown days30 / 0
Training / display graindaily / daily
Horizon / intervals14 days / 95% and 80%
Engine / growth / seasonality modeProphet 1.1.7 / linear / additive
Daily / weekly / yearly seasonalityOff / On / Off
Changepoint / seasonality / holiday prior scales0.05 / 1 / 0.1
Changepoints / predictive draws / seed5 / 1000 / 42
Known events / MCMC samplesNo event calendar applied / 0
Prediction floor$0.00
Holdout window (UTC)2026-03-24 → 2026-03-30
Holdout MAPE30.33% (7 nonzero days)
95% / 80% holdout coverage42.86% / 28.57%

13 rows. Scroll inside the table to see more.

Compare backtest candidates
AWS final-week candidate comparison; costs in USD
CandidateRMSEMAPE95% coverage80% coverage
daily, flat, weekly off$77.4563.74%28.57%28.57%
daily, linear, weekly on$40.7930.33%42.86%28.57%
hourly, linear, weekly on$40.9632.03%28.57%0%

3 rows. Scroll inside the table to see more.

How the forecast flags anomalies

Each day has an expected cost and an expected range. When observed cost rises above the 95% range, the forecast flags a spike. When it falls below the range, it flags a drop. Coral markers locate those days on the chart; the table gives the actual cost, expected cost and deviation.

The 95% band is wider and produces fewer alerts. The 80% band is tighter and brings changes into view earlier. Severity measures the distance beyond the 95% band divided by that band’s width: a value of 1 means the point is one full band width past the boundary.

Use spikes to investigate runaway resources and unexpected billing events. Use drops to check failed jobs, stopped workloads and lost data feeds. Compare each flag with deployment records and known business events to decide what needs action.

Score new data daily against the forecast band. Alert on consecutive breaches, route the alert to the service owner and re-fit regularly. Add known holidays or planned events as model inputs; use regressors for predictable changes in workload so the team can focus on unexplained movement.