←back to features

Agentic SQL

At ChatBotKit, we believe AI agents should have seamless access to data - regardless of where that data lives or how it's normally accessed. That's why we've developed Agentic SQL, a powerful capability that lets your AI agents use SQL commands to query third-party APIs, cloud storage, filesystems, and other data sources, even when those services don't natively support SQL.

Traditional integrations often require your AI agents to learn dozens of different API operations, each with their own syntax and quirks. Agentic SQL changes this by providing a unified, familiar query language that developers and AI agents alike can use to retrieve, filter, and aggregate data across diverse systems.

How It Works

Agentic SQL implements intelligent SQL endpoints that act as proxies between your AI agents and external data sources. When your agent issues a SQL query, ChatBotKit translates it into the appropriate native operations for each connected service—whether that's a REST API call, a database query, or a cloud storage operation.

This translation layer handles all the complexity behind the scenes:

  • Query Parsing: Your SQL query is parsed and analyzed to understand the data requirements
  • Operation Mapping: The query is converted into one or more native API calls
  • Result Aggregation: Responses from multiple operations are combined into a unified result set
  • Filtering and Sorting: Complex WHERE clauses and ORDER BY operations are applied consistently

The result is a consistent, predictable interface that your AI agents can rely on, regardless of the underlying data source.

Key Benefits for Agentic AI

Agentic SQL is specifically designed to support advanced AI agent workflows:

Query Discoverability

AI agents can explore data structures using familiar SQL patterns. Instead of parsing complex API documentation, agents can use standard SQL introspection to understand what data is available and how to access it.

Multi-Step Reasoning

Complex AI tasks often require chaining multiple data operations together. With Agentic SQL, agents can compose sophisticated queries that span multiple data sources, enabling them to gather context, make decisions, and take actions in coordinated sequences.

Data Aggregation

When AI agents need to summarize large datasets or compute statistics across records, Agentic SQL provides native support for aggregation functions. This is far more reliable than having the language model perform calculations directly.

Operational Consistency

Whether your agent is querying a CRM, analyzing files in cloud storage, or pulling records from a database, the SQL interface remains consistent. This reduces the cognitive load on your AI systems and minimizes errors.

Supported Data Sources

Agentic SQL currently supports a growing list of services and data sources:

CRM and Business Systems

  • HubSpot CRM (contacts, companies, deals, and all objects)
  • Additional CRM integrations coming soon

Databases

  • Supabase for PostgreSQL databases
  • Direct database connections for supported providers

File Storage

  • Query and analyze CSV, Excel, JSON files from various sources
  • Cloud storage integrations for file-based data analysis

API Services

  • REST API proxies that expose SQL interfaces over standard HTTP APIs
  • Custom integrations through the ChatBotKit ability system

Example Query Patterns

Here are some examples of how Agentic SQL simplifies data access:

Querying CRM Data

Aggregating Records

Cross-Source Analysis

Extensibility and Future Development

Agentic SQL is designed for extensibility. The architecture supports adding new data sources and services through the ChatBotKit ability system. As we expand support for additional integrations, your existing SQL queries will continue to work—you'll simply gain access to new data sources using the same familiar syntax.

We're actively working on expanding Agentic SQL capabilities, including support for more CRM platforms, additional cloud storage providers, and enhanced analytical functions tailored for AI agent workflows.

Getting Started

Ready to give your AI agents SQL superpowers? Agentic SQL is available through ChatBotKit's skillset and ability system. Configure the appropriate SQL ability for your target data source, connect your credentials, and your agents will be ready to start querying.

For HubSpot integration, explore the HubSpot abilities in the ChatBotKit ability catalogue. For database queries, check out the Supabase SQL ability or connect directly to supported database providers.

Agentic SQL transforms how your AI agents interact with data, providing a powerful yet familiar interface that simplifies complex operations and enables sophisticated multi-step reasoning across diverse data sources.

agenticsqlaidata