When the model qualifies
Deploy the evaluated version to your project’s API. Try predictions in the playground, connect from your server, and route low-confidence results to review.
How it works
From your examples.
Into your product.
Bring your labeled data and describe the task. SotaYard handles preparation, training, evaluation, and deployment. Follow each step in one workspace.
Start your first taskSupport routingIllustration
01 / Your dataset
A few examples.
A clear task.
support-tickets.csvInput: message · Answer: label
CSV| Message | Label |
|---|---|
| Please resend my invoice | billing |
| The export keeps failing | technical |
| I was charged twice | billing |
You supply the answers. The model learns the pattern.
Support routingIllustration
02 / Managed training
One task.
A model of its own.
- 01Check your examplesLabels, duplicates, and class balance
- 02Fine-tune the modelPretrained encoder → your labels
- 03Evaluate a candidateReserved examples, independent of training
Follow each run from preparation through evaluation.
Support routingIllustration
03 / Evaluation
Evidence comes
before deployment.
TrainSelectCalibrateTest
- Accuracy
- How often accepted answers are right
- Coverage
- How much work the model can accept
- Baseline
- A reference to compare the model against
Only a model that qualifies can be deployed.
Support routingIllustration
04 / Your model, available by API
Ready for the
next real example.
POST
/api/v1/projects/:id/predict{
"inputs": {
"message": "Please send my invoice"
}
}Example labelbilling
Test in the playground. Connect with an API key.
Before you start
A clear task.
Examples to learn from.
No training code to write. You bring the task, examples, and quality requirements.
Read the data and API guide ↗01 / The task
One decision, clear labels.
Describe what the model should classify and what each label means. The beta supports text and structured records with 2–12 possible labels.
02 / The examples
Inputs with the right answers.
Bring CSV or JSONL with labeled examples. The beta accepts up to 2,000 rows and 5 MB per dataset. Include a group ID for related examples.
03 / The requirements
Decide what good enough means.
Set your accepted-accuracy and automatic-coverage requirements before training. Your model must qualify against these requirements before deployment.
After training
Your results guide
the next step.
When it needs more work
Read the evaluation and its qualification reasons. Improve your examples, add genuinely new data, and train another version. A finished run is not automatically deployed.
Give your data
a job to do.
Start with your own labeled examples, or try the sample dataset. Train, evaluate, and serve a text classification model in one workspace.
Start building free