01
Project
Choose or create your Forecasting project.
A scope for saved data, model versions and forecast history
- Scope
- Forecasting project
- History
- Saved models and runs
- Selection
- Choose intentionally
MAG OptiAIOperational AIForecastingEarly Access
Plan What Comes Next with Forecasts You Can Review.
Turn historical data into forecasts you can inspect. Prepare your series, train and compare saved models, then review the forecast range and planning context before deciding what to do next.
Prepare the history. Choose a saved model. Review the forecast before putting it into a plan.
Prefer to explore first? Start free14-day trial · lower limits. Beta access requires approval.
Built for planning teams
Historical data and forecasts
Cloud-based Pro Console
Guided evaluation
The forecasting challenge
A Forecast Has to Mean More Than a Predicted Number.
Historical data, model choice and future assumptions shape the result you use for planning.
Challenge 01 / 05
A Forecast Starts with the History You Supply
Missing or inconsistent data can undermine the planning question.
Dates, values, frequency and optional drivers need to describe the series you want to forecast. Preparing the data is part of the workflow, rather than an invisible step behind the result.
Impact
Review input readiness before training a saved model.
The challenge is more than drawing a future trend.
It is preparing and reviewing a forecast with its data, model, assumptions and evaluation context connected.
Why MAG OptiAI
A Clearer Way to Build and Review Forecasts.
Keep the data, saved model choice and forecast execution context connected throughout the workflow.
A Clear Data Foundation
Prepare or restore a named data draft with the historical series and optional drivers for your planning question.
Saved Models with a Deliberate Choice
Train candidates, inspect saved versions and select or stage the model used for execution.
Forecasts with Review Context
Inspect charts, returned rows, model provenance and forecast ranges before using an output for planning.
Your forecasting context stays connected.
Saved projects hold data drafts, model versions and forecast runs. Review which model and inputs produced the output you are using.
Process
How We Work
From historical data to a forecast you can review.
MAG OptiAI guides your team from preparing the forecasting data to selecting a model and reviewing its forecast.
02
Data
Prepare and validate the forecasting data draft.
Historical series · frequency · optional drivers · upload or samples
- History
- Dates and values
- Drivers
- Where applicable
- Readiness
- Review input checks
03
Models
Train candidates, compare versions and choose a model.
SARIMA · ETS · LightGBM · saved versions · model staging
- Train
- Saved candidates
- Compare
- Compatible versions
- Stage
- Default model choice
04
Forecast
Choose the execution source and generate a forecast.
Staged model or selected saved version · horizon · applicable future drivers
- Source
- Staged or selected version
- Horizon
- Supported number of steps
- Drivers
- Matching future inputs
05
Results
Review the output before putting it into a plan.
Chart · forecast rows · range · provenance · exports · AI interpretation
- Review
- Forecast and model context
- History
- Open a saved run
- Use
- Export for planning review
Core capabilities
What Forecasting Can Do for Your Organization
A connected workflow for preparing data, choosing saved models and inspecting forecasts.
Prepare Historical Data
Upload or use samples, inspect readiness and restore saved data drafts.
Train Saved Model Candidates
Use supported SARIMA, ETS and LightGBM model families.
Compare Model Versions
Review compatible saved candidates using available evaluation context.
Patterns into forecasts.
Decisions with context.
Choose the Execution Source
Forecast with the staged model or a ready selected saved version.
Supply Applicable Future Drivers
Match future inputs to the saved model's required feature schema.
Inspect and Export Results
Review the chart, returned rows, forecast range and saved run context.
Stage a Default Model
Make a deliberate saved-model choice for future forecasts.
Use Grounded AI Interpretation
Ask for interpretation tied to the successful forecast and its available context.
Prepare Historical Data
Upload or use samples, inspect readiness and restore saved data drafts.
Train Saved Model Candidates
Use supported SARIMA, ETS and LightGBM model families.
Compare Model Versions
Review compatible saved candidates using available evaluation context.
Stage a Default Model
Make a deliberate saved-model choice for future forecasts.
Choose the Execution Source
Forecast with the staged model or a ready selected saved version.
Supply Applicable Future Drivers
Match future inputs to the saved model's required feature schema.
Inspect and Export Results
Review the chart, returned rows, forecast range and saved run context.
Use Grounded AI Interpretation
Ask for interpretation tied to the successful forecast and its available context.
All in one place.
The tools to prepare, forecast and review in one connected workflow.
Forecasting in action
See the Context Behind the Forecast Line.
Review historical values, returned forecast points and the range associated with the model. This illustration explains the concept; it is not a measured model result.
History, horizon and range belong in the same review.
Inspect the forecast chart alongside the returned rows, warnings and saved-model execution context. The range's meaning depends on the model; it is not a promise that every future value will fall inside it.
Evaluate the forecast against your planning question and the available evaluation information.
History and forecast horizonIllustrative
Historical valuesForecastIllustrative range
| Period | Forecast | Illustrative range |
|---|---|---|
| Future 1 | 104 | 99–110 |
| Future 2 | 107 | 100–115 |
| Future 3 | 105 | 96–117 |
Who it’s for
Is Forecasting a Good Fit for Your Team?
For teams using historical series to inform a defined planning question.
Demand and Operations Planners
Prepare a relevant series, frequency and forecast horizon for planning demand, volumes or activity.
Analysts Comparing Models
Train saved candidates and inspect compatible versions before choosing the execution model.
Teams Reviewing Forecast Outputs
Examine the forecast, its range, assumptions and available evaluation context before acting.
Every series and planning question is different. Evaluate the supported models and inputs against your requirements; historical patterns do not guarantee future outcomes.
The Forecasting Evaluation
Real Problem. Real Evaluation. Real Impact.
Bring your forecasting question and evaluate the fit with your organization’s information.
What You Get
Access for a Larger Evaluation
Apply through the existing beta process to evaluate beyond the smaller free-trial limits.
A Guided Data Setup
Frame the planning question and prepare history, frequency and applicable driver inputs.
A Saved-Model Workflow
Train model candidates and inspect the versions used for your forecasts.
A Direct Feedback Channel
Discuss setup difficulties, unclear results and requirements that need a better fit.
Results You Can Review
Inspect outputs and available evaluation context to judge suitability for your planning needs.
What We Ask From You
A Defined Forecasting Question
Bring the series, frequency, forecast horizon and decision your team wants to inform.
Usable Historical Data
Provide dated values and any relevant drivers. Anonymized data is welcome when it preserves the pattern and required relationships.
Practical Feedback
Tell us where preparation, model selection or result interpretation needs improvement.
A Review of the Outcome
Check the forecast, assumptions and evaluation context before relying on the output.
Real Problem
Real Evaluation
Better planning decisions.
A real-world evaluation with your organization.
Bring your forecasting problem and real or anonymized historical data. We will review the fit and help you get started.
Forecasting Beta
Frequently Asked Questions
What to know before evaluating your forecasting requirements.
What data do I need?
Start with a dated historical series and a clear forecasting question. Prepare the frequency and optional drivers required by your chosen model, then review the product's input requirements and readiness checks.
Which model families are available?
The current workspace supports SARIMA, ETS and LightGBM. Appropriate settings and input requirements depend on the family and the dataset.
Does training automatically choose the model?
Training creates saved candidates. Staging deliberately selects a default model. Forecast execution can also target a ready selected saved version without changing that default.
Can models use external drivers?
Supported models can use driver inputs where applicable. Future values need to match the saved model's feature schema and execution requirements; review which values were supplied or filled in the execution provenance.
What does the forecast range mean?
The displayed range depends on the model and saved execution metadata. LightGBM can return an empirical range, while other supported families use their native forecast ranges. Review that context rather than assuming every shaded band has identical statistical meaning.
Does this guarantee accurate predictions?
No. Forecast usefulness depends on the history, assumptions, model and evaluation context. Inspect the results and warnings and assess their suitability for your planning decision.
How does Early Access relate to the trial and beta?
Early Access describes product maturity. Start Free uses the existing 14-day trial with lower limits. Larger evaluations use the existing beta application and team approval; there is no separate access-request system.
What can AI interpretation explain?
AI interpretation can discuss the successful forecast's direction, range, model family, available evaluation context and warnings. It does not establish causality or guarantee future outcomes.
How do I apply for a larger evaluation?
Use Apply for the Forecasting Beta to reach the existing product-specific application. Our team reviews your requirements and confirms the fit for an evaluation.
Still have a question?
Talk to us about your forecasting evaluation.
Ready to get started?
Ready to Evaluate Forecasting with Your Organization’s Data?
Tell us about your historical series, forecast horizon and planning requirements. We will review your application and confirm whether the current Forecasting Beta is a good fit for your organization.
No payment requiredApplication reviewed by our team