Sema4.ai Data Access

Sema4.ai Data Access provides enterprise agents a zero-copy data access to past, present, and future data. Read more about the offering at our product site.

This VS Code extension provides the core functionality for developing Sema4.ai Actions with data access. The extension is designed to work in tandem with the Sema4.ai SDK extension for VS Code. Download it here before getting started.

Features

• Runs and manages the state of Sema4.ai Data Server that provides access to 100s of data sources and predictive models.
• Add, edit and delete data sources (like Postgres, Snowflake, files and more).
• Provides an explorer view to your current Data Server state.
• Run any SQL queries in .sql files in your project to develop and iterate queries that your agent can access.
• Create named queries in your Sema4.ai Actions project with an intuitive UI that lets you choose the parameters you want, and easily write the necessary documentation string.
• Create and test powerful predictive models and knowledge pages to your Data Server.
• View query history.

Getting started

1. Install the extension, along with the Sema4.ai SDK VS Extension.
2. Create a new Sema4.ai Action project from a templates from Command Palette (Cmd-Shift-P) Sema4.ai Create Action Package. Follow the steps and choose any of the Data Access templates.
3. At first launch, the Data Server will download and bootstrap. You will see the status indication in status bar at the bottom of the screen, with green "Data Server is running" indication when everything is ready.
4. You are good to go!

Usage

1. The extension maintains the state of Sema4.ai Data Server, and you can stop and start the server.
2. View the data sources, models and files in your Data Server in the explorer view. Click any of them to make a quick query. Use Command Palette to Add New Data Source that connects to any external database, and Upload File to add file data sources.
3. Open any .sql file and write queries. You are then able to run them against your data sources and see the results in a panel that appears on the side.
4. When happy with the queries, create Sema4.ai Named Queries that allow agents to access your data in secure and scalable way.

When creating a named query, you'll be asked to write the name (of the Python function), the docstrings and choose which parts of the query to make parameters of your named query.

Once saved, the extension will create a data source definition in your project, add a named query python code and a test data entry.

Learn more

Head over to Sema4.ai Documenentation to learn more about building Actions with data access, and using them with your agents.

Changelog

View the latest changes in this VS Code extension through the CHANGELOG.

Credits

The extension is based on the great work by GitHub user cweijan and his vscode-database-client. Thank you!