Skip to content
RelationalAI Docs

Build With RelationalAI

Model how your business works with PyRel, RelationalAI’s Python SDK. Define its concepts, relationships, and logic in a semantic model connected to your Snowflake data. Applications can query the model instead of reimplementing business logic over source data. Connected AI agents can use the same model to answer questions in your business’s terms.

Where to start

Section titled “Where to start”

How PyRel works

Section titled “How PyRel works”

Use results in

  • Applications
  • BI
  • Agents
Your development environment
PyRelDeclarative definition of your semantic model in Python
  • Define the kinds of things your business tracks, such as customers and orders, and the properties that describe them.
  • Define how concepts connect, such as which customer placed an order.
  • Use rules to derive new facts and check whether data meets business requirements.
  • Use your model to see how things are connected, choose the best option for a goal, or predict an outcome.
Send model and query
Return results to PyRel
Use PyRel to model business concepts and relationships, express rules, and reason over Snowflake data to answer business questions: how entities are connected, which decisions best meet your goals, or what outcomes are likely. The Native App evaluates submitted work in your Snowflake account and returns results to Python for use in applications, BI, or agents.

Explore the documentation

Section titled “Explore the documentation”