Introducing Inferable Data Connector - agentic automation for your Databases and APIs

A secure bridge that enables natural language queries for databases and APIs while maintaining data privacy and security. Features PostgreSQL, MySQL, SQLite, OpenAPI, and GraphQL support.

Nadeesha Cabral on 08-12-2024

Inferable Data Connector is a secure bridge that connects your data systems to Inferable while keeping your sensitive data private and secure. This connector enables natural language interactions with your data without exposing your systems to the public internet.

Data Connector

We talked about how we use text to sql to automate some of our internal workflows at Inferable. Data Connector is the next evolution of this approach.

We've also put together a quick demo of using Postgres and OpenAPI with real life scenarios how we use them.

Key Features

  • 🔐 Secure by Design: Your credentials and sensitive data never leave your environment
  • 🌐 Zero Inbound Access: Runs inside your network and pulls instructions from Inferable
  • 🧩 Highly Extensible: Add new data sources by simply writing a new function
  • 🔄 Schema-Aware: Automatically adapts to schema changes in your data systems
  • 🤿 Privacy-First: Optional privacy mode ensures query outputs stay private (not sent to the model)
  • 🔍 Added Security: Optional approval mode for query execution control

Supported Data Sources

The connector launches with support for:

  • PostgreSQL
  • MySQL
  • SQLite
  • OpenAPI (Experimental)
  • GraphQL (Experimental)

Privacy and Security

We understand that data security is paramount. The Data Connector includes:

  • Privacy Mode: Raw data never leaves your environment - only schema information is shared
  • Approval Mode: Optional manual approval for query execution
  • Local Execution: All queries run within your environment
  • Read-Only Access: Option to use read-only database connections

What's Next?

We're actively working on:

  • Supporting more data sources
  • Enhanced security features
  • Improved performance for large datasets
  • Better schema handling for complex data structures

Try It Today

The Data Connector is available now as part of both our cloud offering and self-hosted installations. Check out the documentation to get started, or join the community to share your experience and contribute to its development.

Visit the Inferable GitHub repository to explore the code and get started with the Data Connector today!

What can I use it for?

Here are some real-world scenarios where the Data Connector shines:

Customer Support Automation

Imagine you're a Support Lead at a SaaS company managing customer inquiries. Daily, you need to:

  • Look up customer details in your Postgres database
  • Check subscription status in Stripe
  • View support ticket history in Zendesk
  • Verify feature usage in your analytics database

Instead of jumping between tools, you can create a workflow that pulls this information with a single query like "show me all details for customer support@example.com including their subscription status and recent tickets".

Developer Operations

As a DevOps engineer at a startup who has to manage multiple systems, you frequently need to:

  • Check GitHub PR status across multiple repositories
  • Verify deployment logs in your internal Postgres database
  • Monitor system health metrics in SQLite-based monitoring
  • Track API usage patterns

Create a workflow that lets you ask "show me all failed deployments from last week and their related PRs" or "which APIs are experiencing higher than normal error rates?".

Sales Pipeline Management

As a Sales Operations Manager using HubSpot and Salesforce, you regularly:

  • Sync contact data between platforms
  • Update deal stages across systems
  • Generate pipeline reports combining multiple data sources
  • Track sales team performance metrics

Build a workflow to ask "show me all deals above $50k that haven't been updated in 7 days" or "generate a report of all leads from last month's campaign with their current status".

Product Analytics

As a Product Manager working with multiple data sources, you need to:

  • Query user behavior from your Postgres database
  • Check feature adoption rates
  • Monitor subscription changes in Stripe
  • Track user feedback from Zendesk

Create workflows to ask "show me all users who tried feature X but didn't convert" or "what's the adoption rate of our new feature among enterprise customers?".

Engineering Debugging

As a Backend Engineer debugging production issues, you often need to:

  • Query logs from your production Postgres database
  • Check API responses from internal services
  • Verify webhook deliveries in Stripe
  • Look up user actions in your audit logs

Set up a workflow to ask "show me all failed transactions for user X in the last hour across all systems" or "what was the API response when this webhook failed?".

Open Source and Self-Hostable

The Data Connector is open source and self-hostable. You can run it on your own infrastructure alongside Inferable.

Subscribe to our newsletter for high signal updates from the cross section of AI agents, LLMs, and distributed systems.

Maximum one email per week.

Subscribe to Newsletter