Skip to main content
Lyzr Agent Studio allows your AI agents to connect seamlessly with all leading SQL (Structured Query Language) and NoSQL data sources available in the market. This critical capability enables your agents to move beyond static, pre-loaded information and access, analyze, and reason over your organization’s real, live operational data in real time. This integration is the foundation for creating powerful Data Query Agents, which are often referred to as Text-to-SQL Agents. Once connected, your agent can:
  1. Receive natural language questions from a user (e.g., “Show me last month’s sales by region”).
  2. Convert that natural language question into an accurate, executable database query (such as a complex SQL statement).
  3. Execute the query against the database.
  4. Process the results and provide accurate, data-driven insights in clear, natural language back to the user in real time.
🎥 Here’s a quick video on how to build a Text-to-SQL Agent: Watch Video

The Role of the Semantic Model

Connecting an entire, complex, and often massive corporate database directly to a single agent is inefficient, resource-intensive, and provides excessive, irrelevant context. Lyzr solves this challenge by requiring the creation of a Semantic Model.
  • Definition: A Semantic Model is a crucial conceptual layer that holds only a relevant subset of your database schema. It includes only the specific tables, columns, and relationships required for that particular agent’s task.
  • Purpose: By holding only the relevant information, the Semantic Model ensures:
    • Optimal Performance: The agent has a focused scope, which drastically improves the speed and accuracy of query generation and execution.
    • Accurate Context: It gives the agent the precise, necessary context to understand your data relationships and terminology, allowing it to generate highly accurate and context-appropriate queries.
    • Data Governance: It acts as a controlled boundary, ensuring the agent can only access the data authorized for its function, enhancing security.

🔌 Supported Database Connectors

Lyzr offers robust, built-in connectivity to a comprehensive range of popular and enterprise-grade data stores, covering major cloud and on-premise solutions. The image below displays the variety of database connectors available within the Lyzr Agent Studio:
TypeData ConnectorDetailsStatus
Relational (SQL)PostgreSQLA powerful open-source object-relational database system.Fully Supported
MySQLThe world’s most popular open-source relational database.Fully Supported
Amazon RedshiftAmazon’s fully managed, petabyte-scale cloud data warehouse service.Fully Supported
Google BigQueryGoogle Cloud’s serverless, highly scalable, and cost-effective cloud data warehouse.Fully Supported
Azure SQLMicrosoft’s intelligent, scalable, cloud database service.Fully Supported
Microsoft SQL ServerThe widely-used enterprise-level relational database management system.Fully Supported
SnowflakeThe cloud-based data platform known for its unique architecture and scalability.Upcoming
NoSQLMongoDBA leading general purpose, document-based distributed database.Fully Supported
Cloud/Data LakeDatabricksIntegration with the unified data platform for data engineering, ML, and analytics.Upcoming
FilesFile Upload (CSV, Excel, JSON)Direct connectivity for analyzing data stored in common flat file formats.Fully Supported