PowerMind Predict Overview
PowerMind Predict is the forecasting app in PowerMind AI Studio. You upload time-series CSV data (or connect stored uploads), run a prediction job (Prophet / ARIMA / XGBoost), then review forecasts, MSE/RMSE, and audit history.
It lives under AIAAS in the Studio marketplace, next to RAGBot, DocRead, Transcribe, and VisionID.
What you see in the product
After you open Predict, five tabs run left to right:
| Tab | What it is |
|---|---|
| Dashboard | Live overview: total / success / failed jobs, users, latest MSE/RMSE, recent jobs |
| Data Sources | CSV uploads that feed prediction jobs (rows, columns, workspace, created date) |
| Prediction Jobs | Create and track train/forecast jobs (+ New prediction) |
| Predictions | Result view for a job: status, model, periods, MSE/RMSE, chart & forecast table / export |
| Audit Logs | Who logged in or changed data — timestamp, actor, action, target, IP |
How it works (simple)
flowchart LR
A[Sign in to Studio] --> B[Open Predict]
B --> C[Upload CSV / Data Sources]
C --> D[New prediction job]
D --> E[Wait for SUCCESS]
E --> F[Predictions / export]- A CSV with a date column and a numeric target is uploaded.
- Predict trains a model (for example Prophet) and forecasts N periods.
- Quality is shown as MSE and RMSE (lower is better — not a % accuracy score).
- You can save the model later via API, or re-run with a loaded model.
Always check the project in the top bar (example: AI Project) so you are in the right workspace.
Where to go next
- Getting Started — first login to first forecast
- Web UI — every screen with screenshots
- How it works — the full day-to-day flow