# Pascal Models

Source: https://www.trypascal.io/platform/models

A better score is only the beginning.

Connect prediction objectives, evaluation results, and release decisions. Give each model version a clear case for further study or use in the operation.

## Capabilities

### Start with the decision

Specify what to predict, for whom, how far ahead, and what acceptable performance means. Keep the target, risk, and review requirements attached to the objective revision.

### Fix the data behind each run

Connect training datasets and feature releases to the exact inputs used. Review freshness and coverage before a training or prediction command proceeds.

### Compare the tradeoffs

Inspect training trials, pin an explicit baseline, and compare measured quality and operational metrics. Keep the decision to select, reject, or continue separate from the comparison view.

### Evaluate before release

Review results on a fixed sample, including relevant slices, guardrails, and insufficient-data outcomes. Carry the evaluation into the review of a model version.

### Deliver an exact version

Publish an immutable model release for batch or online delivery. Follow deployment, canary, promotion, and rollback through the serving controls.

### Keep learning from the operation

Connect observations and labels to monitoring. Investigate incidents and propose evaluation, annotation, rollback, or retraining with an authorized next action.

## Connected context

Connect data and features to training candidates and evaluation results. Model Registry and Serving handle releases and predictions; Lab and monitoring bring feedback into the next review.

- Modeling objectives, revision history, targets, metrics, and guardrails
- Dataset and feature releases, training jobs, trials, and evaluation results
- Model releases, batch predictions, deployments, monitor windows, and feedback

## Controls and permissions

Each step retains its own inputs, review, and result. The walkthrough demonstrates candidate comparison and a decision to continue studying. Hosted language-model providers and routing are managed separately in AI Models.

## Example

Meridian compares three candidates for predicting which shipments will miss the 21:00 cutoff. Trial 3 improves the overall score, but hazardous-load recall falls below the 90% minimum. Maya records Continue study with a reason to improve recall before proposing a release.

## Workflow

### Define the objective

- Keep the decision, training inputs and guardrails together. Fix the criteria that will make a prediction useful to the operation.

### Compare candidates

- Inspect the baseline, evaluation results and slice performance. Record whether to select, reject or continue studying the candidate.

### Plan its operating life

- Take an evaluated version through release and the qualified delivery path. Monitor the result against the decision it is intended to support.

## Next steps

Bring one model objective, its baseline, and the guardrails that matter in your operation. Define what a candidate must demonstrate before release.

- [Contact sales](https://www.trypascal.io/contact)
- [Logistics operating system](https://www.trypascal.io/solutions/logistics)