Working with Models

Models represent AI/ML models that are part of your use cases in Credo AI.

Overview

Models in Credo AI represent the machine learning models you're building, training, or deploying. Each model can be associated with one or more use cases.

Listing Models

Python

from credoai import CredoAI

client = CredoAI()

response = client.models.list(page_limit=50)
for model in response.items:
    print(f"{model.name} (ID: {model.id})")

TypeScript

import { createCredoAIClient } from '@credo-ai/sdk';

const client = createCredoAIClient('your-tenant');

const { data: response, error } = await client.models.list({ pageLimit: 50 });
if (error) throw error;
for (const model of response.items) {
  console.log(`${model.name} (ID: ${model.id})`);
}

Creating a Model

Python

from credoai import CredoAI, ModelCreate

client = CredoAI()

model = client.models.create(
    data=ModelCreate(
        name="Churn Predictor v2",
        summary="XGBoost model for customer churn prediction",
    )
)
print(f"Created: {model.id}")

TypeScript

import { createCredoAIClient } from '@credo-ai/sdk';

const client = createCredoAIClient('your-tenant');

const { data: model, error } = await client.models.create({
  name: 'Churn Predictor v2',
  summary: 'XGBoost model for customer churn prediction',
});
if (error) throw error;
console.log(`Created: ${model.id}`);

Getting a Model

Python

model = client.models.get(model_id="model_xyz789")
print(f"Name: {model.name}")

TypeScript

const { data: model, error } = await client.models.get('model_xyz789');
if (error) throw error;
console.log(`Name: ${model.name}`);

Updating a Model

Python

from credoai import ModelUpdate

updated = client.models.update(
    model_id="model_xyz789",
    data=ModelUpdate(summary="Updated summary"),
)

TypeScript

const { data: updated, error } = await client.models.update('model_xyz789', {
  summary: 'Updated summary',
});
if (error) throw error;

Deleting a Model

Python

client.models.delete(model_id="model_xyz789")

TypeScript

const { error } = await client.models.delete('model_xyz789');
if (error) throw error;

Managing Relationships

Models can be linked to use cases and vendors:

Python

from credoai import RelationshipAdd

# Add a model to a use case
client.use_case_models.add(
    use_case_id="uc_abc123",
    data=RelationshipAdd(id="model_xyz789"),
)

# List use cases for a model
use_cases = client.use_case_models.list(use_case_id="uc_abc123")

# Add a vendor to a model
client.model_vendors.add(
    model_id="model_xyz789",
    data=RelationshipAdd(id="vendor_def456"),
)

TypeScript

// Add a model to a use case
const { error: addError } = await client.useCases.models.add('uc_abc123', {
  id: 'model_xyz789',
});
if (addError) throw addError;

// List use cases for a model
const { data: useCases, error: listError } = await client.useCases.models.list('uc_abc123');
if (listError) throw listError;

// Add a vendor to a model
const { error: vendorError } = await client.models.vendors.add('model_xyz789', {
  id: 'vendor_def456',
});
if (vendorError) throw vendorError;

Next Steps

• Learn about Use Cases to organize your AI applications
• Explore Vendors for third-party AI management
• Understand Relationships between resources
• See the full API Reference for all available methods