Skip to main content

Python SDK

Develop new Data Agents using the Heisenberg Python SDK framework.

Overview​

The Heisenberg Python SDK provides a framework for building and productionizing Data Agents. Use it to create custom data processing workflows, define transformers, and build agents that transform raw data into structured intelligence.

For using existing Data Agents, use COOK - the visual interface that allows you to access and use pre-built agents without writing code.

Installation​

Install the latest version of the Heisenberg Python SDK:

pip install sdk-seshat-python

Optional Dependencies​

For Flipside data provider support:

pip install "sdk-seshat-python[flipside_support]"

For PostgreSQL support:

pip install "sdk-seshat-python[postgres_support]"

When to Use the SDK​

Use the Python SDK when you need to:

  • Develop custom Data Agents - Build new agents with specific data processing logic
  • Create custom transformers - Implement domain-specific data transformations
  • Integrate with existing systems - Connect Data Agents to your Python applications and workflows
  • Productionize agents - Deploy and manage agents at scale

When to Use COOK​

Use COOK when you want to:

  • Use existing Data Agents - Access pre-built agents from the marketplace
  • Configure agents visually - Set up agents without writing code
  • Quickly get started - Start using data intelligence without development work

Development Workflow​

  1. Define your data source - Specify where your data comes from
  2. Create transformers - Build or use existing transformers for your processing logic
  3. Configure your Data Agent - Assemble transformers in the correct sequence
  4. Create a FeatureView - Combine your data source and Data Agent
  5. Run and test - Execute your agent and validate the output

Integration with Heisenberg Platform​

Data Agents developed with the SDK integrate seamlessly with the Heisenberg platform:

  • Deployment - Deploy agents to Heisenberg nodes for execution
  • Marketplace - Share your agents in the Data Agent Marketplace
  • Monitoring - Track agent performance and data quality
  • Scaling - Leverage Heisenberg's decentralized infrastructure

Next Steps​