Datasparc Extends Zero Trust Database Access to AI Agents

Datasparc has enhanced its DBHawk Zero Trust platform for database security and access with two new artificial intelligence features: Hawk AI and the Hawk AI MCP Server. Hawk AI serves as a natural-language interface for querying databases, while the Hawk AI MCP Server enables governed and audited access for AI agents to enterprise data. Both solutions operate within the customer's own infrastructure, leveraging DBHawk's existing access controls and security measures.

Hawk AI translates natural language queries into SQL commands, executes them, and presents the results in charts. Unlike external AI services, Hawk AI operates locally and supports the use of customer-provided AI models, ensuring that data and schemas remain within the customer's environment.

All queries generated through Hawk AI adhere to the same object- and column-level access controls and dynamic data masking applied to human users. Every prompt, query, and result is meticulously logged for audit purposes, ensuring that the AI can only access data to which the requesting user is permitted.

"Every company wants the productivity of AI on their data, but not at the cost of a breach or a failed audit. With Hawk AI and the MCP Server, the AI is just another governed user of DBHawk — it only sees what it's allowed to, and everything it does is masked and logged," said Manish Shah, CEO of Datasparc.

The Hawk AI MCP Server implements the Model Context Protocol, allowing any compatible agent to query databases through DBHawk's governed tools. Each agent uses a role-based token, and all requests pass through access control, masking, and read-only restrictions. Actions are audited per agent, maintaining Zero Trust access even for automation. DBHawk supports over 20 databases and is SOC 2 Type II compliant.