Skip to content
Last updated

Personal Connectors Overview

🔨 In Development — This section is still being developed and may change.
Personal Connectors enable your AI assistants to securely connect with external services, APIs, and data sources. This allows your assistants to access real-time information, perform actions on external platforms, and integrate with your personal workflows.

What are Personal Connectors?

Personal Connectors are secure, authenticated connections between your Freddy AI assistants and external services. They allow your AI assistants to:

  • Access real-time data from external APIs
  • Perform actions on connected platforms
  • Integrate with your personal tools and services
  • Maintain security and privacy through token-based authentication

How They Work

1. Configuration

Set up connectors through the Freddy Hub or API:

# Create a connector configuration
configuration = freddy.personal_connectors.configurations.create(
    connector_type="notion",
    name="My Notion Workspace",
    credentials={"api_key": "your_api_key"},
    settings={"workspace_id": "your_workspace_id"}
)

2. Authentication

Connectors use secure authentication methods:

  • API keys and tokens
  • OAuth flows
  • Personal access tokens
  • Custom authentication schemes

3. Function Integration

Once configured, connectors become available as functions in your assistants:

# Assistant can now call connector functions
assistant = freddy.assistants.create(
    name="Connected Assistant",
    tools=[
        {
            "type": "function",
            "function": {
                "name": "notion_search",
                "description": "Search Notion pages",
                "parameters": {...}
            }
        }
    ]
)

Available Connector Types

📝 Productivity Tools

  • Notion - Access workspaces, pages, and databases
  • Google Workspace - Gmail, Google Drive, Calendar
  • Microsoft 365 - Outlook, OneDrive, Teams
  • Slack - Send messages, access channels
  • Discord - Interact with servers and channels

🗄️ Databases & APIs

  • PostgreSQL - Query databases directly
  • MongoDB - Access document collections
  • REST APIs - Connect to any REST API
  • GraphQL - Query GraphQL endpoints

🛠️ Developer Tools

  • GitHub - Access repositories, issues, PRs
  • GitLab - Repository and CI/CD integration
  • Jira - Project management and ticketing
  • Linear - Issue tracking and project management

📊 Business Tools

  • Salesforce - CRM data and operations
  • HubSpot - Marketing and sales automation
  • Stripe - Payment processing and billing
  • Shopify - E-commerce platform integration

Security & Privacy

🔐 Security Features

  • End-to-end encryption for all data transmission
  • Secure token storage with automatic rotation
  • Granular permission controls
  • Audit logging for all connector activities

🔒 Privacy Protection

  • No permanent storage of your credentials
  • Data processed in-memory only
  • Configurable data retention policies
  • GDPR and SOC 2 compliance

Getting Started

Prerequisites

  1. Freddy API account with personal connectors enabled
  2. Valid credentials for the services you want to connect
  3. Understanding of the target service's API

Step 1: Choose Your Connectors

Identify which services you want your assistant to access and ensure you have the necessary permissions and API keys.

Step 2: Create Configurations

Set up connector configurations through the Hub or API:

configurations = freddy.personal_connectors.configurations.list()
for config in configurations.data:
    print(f"Connector: {config.name} ({config.connector_type})")

Step 3: Build Connected Assistants

Create assistants that use your configured connectors:

assistant = freddy.assistants.create(
    name="My Connected Assistant",
    instructions="You can access my Notion workspace and Gmail...",
    tools=[...]  # Connector functions will be available
)

Best Practices

🔑 Credential Management

  • Use environment variables for sensitive credentials
  • Rotate API keys regularly
  • Use the least-privileged access possible
  • Never commit credentials to code repositories

⚡ Performance Optimization

  • Cache frequently accessed data when possible
  • Use pagination for large datasets
  • Implement retry logic for transient failures
  • Monitor connector usage and costs

🛡️ Error Handling

  • Implement graceful fallbacks for connector failures
  • Provide clear error messages to users
  • Log connector activities for debugging
  • Handle rate limits and API quotas

Troubleshooting

Common Issues

Authentication Failures

  • Verify API keys are valid and not expired
  • Check that the service account has necessary permissions
  • Ensure correct authentication method is configured

Connection Timeouts

  • Check network connectivity
  • Verify service availability
  • Implement retry logic with exponential backoff

Permission Errors

  • Review and update connector permissions
  • Check if the connected account has access to requested resources
  • Verify OAuth scopes are correctly configured

Getting Help

  • Check connector-specific documentation
  • Review API logs in Freddy Hub
  • Contact support for connector-specific issues

Next Steps

For technical support, contact support@aitronos.com.