Forecasting.
Time-series forecasting as a service — demand, capacity, load — with uncertainty intervals and what-if scenarios.
Probabilistic forecasts for demand, capacity and load, with covariates, holidays and uncertainty intervals — on the model that already forward-rolls systems, so scenarios are physics-aware, not just curve-fits.
- Category
- Applied AI
- Deployment
- Managed → air-gapped
- Governance
- IAM · encryption · audit
Forecasting libraries are everywhere. Physics-aware scenarios aren't.
Time-series forecasting is a mature field, and a managed service that ingests history, handles covariates and returns probabilistic forecasts is genuinely useful — Aether provides exactly that, with intervals rather than a single fragile point estimate. What's different is the model underneath: because Aether forward-rolls systems, a 'what-if' isn't only a statistical extrapolation, it can be grounded in how the system actually behaves.
Probabilistic, with the drivers that matter.
Forecasts come with uncertainty intervals so you can plan for risk, not just a number. Covariates, prices, weather and calendars enter as drivers, holidays and seasonality are handled, and points that break the pattern are flagged as anomalies rather than silently smoothed.
What-if branching, on the same platform as your data.
Branch a forecast on a decision — a price change, a capacity addition, a demand shock — and compare outcomes before committing. Because it runs on the platform with your warehouse, the history it learns from and the operations it informs share one governed source, with no export to a separate ML service.
Forecast demand with intervals, then branch on a decision — all against the warehouse, no export.
const base = await aether.forecast({
series: "lake.daily_demand", horizon: "90d",
covariates: ["price", "holiday"], intervals: [0.1, 0.9],
});
const whatIf = await base.scenario({ price: "-10%" }); // branch the forecast
compare(base.p50, whatIf.p50);You get a forecast with honest intervals and a what-if branch on a real decision, learned from and informing the same governed warehouse. No copy to a separate forecasting service and back.
Three steps to running.
Point at your time-series in the data platform.
Covariates, prices, weather and calendars shape the forecast.
Probabilistic forecasts with what-if scenarios and anomaly flags.
Forecasting, in full.
Intervals, not just a point estimate, so you can plan for risk.
Prices, weather and holidays as drivers.
Branch the forecast on a decision and compare.
Surface points that break the pattern.
Provision it in a few lines.
Every service is reachable from the same SDK, CLI and infrastructure-as-code — one identity, one bill, one audit trail across the whole catalog.
# Provision forecasting on Aether Cloud
aether applied-ai create \
--service forecasting \
--name app \
--region us-1 \
--deploy managed # or vpc | on-prem | air-gappedAt a glance.
- Output
- Probabilistic intervals
- Drivers
- Covariates · calendars
- Scenarios
- What-if branching
- threat-intel platforms (TIP)es
- Flagged
- Horizon
- Configurable
Built for real work.
Demand & inventory planning
Capacity planning
Financial forecasting
On one model, not stitched together.
The usual stack runs forecasting in one product, the model in another and the data in a third — and the seams between them are the cost. Aether Cloud runs it on the same platform that serves the model, governs your identity and deploys into your boundary, with the rest of the catalog one hop away.
No stitching a vector DB to one place, a warehouse to another and a model to a third — forecasting sits next to the rest of the catalog, one identity, one bill.
The provider that runs Aether runs your forecasting — so the data and the model never leave the same governed boundary to talk to each other.
Need a capability that isn’t here yet? The model writes and deploys it into the same boundary — the catalog is a starting point, not a ceiling.
Good to know.
Probabilistic time-series forecasting — on the model that forward-rolls systems, so scenarios can be physics-aware, not just statistical.
Yes — intervals, not just a point estimate, so you can plan for risk.
Yes — branch the forecast on a decision and compare outcomes.
Your boundary, your choice.
Pairs well with.
Detect, classify and read images and video through one API — grounded in the multimodal model, not a fixed label set.
Speech-to-text and text-to-speech in one service — transcription, diarization and natural voices, on the model that reasons over what's said.
Real-time and batch translation across 100+ languages, with glossaries and the context the model carries — not phrase-book substitution.
Turn documents into structured data — forms, tables, invoices, contracts — grounded in the model, not brittle templates.
Real-time recommendations and next-best-action from your interaction data, served at low latency.
Run Forecasting on Aether Cloud.
Time-series forecasting as a service — demand, capacity, load — with uncertainty intervals and what-if scenarios. Deployable managed, in your VPC, on-prem or fully air-gapped — talk to us about the configuration your workloads and your boundary require.