In the rapidly evolving digital landscape of July 2026, finding intuitive ways to interact with complex databases remains a top priority for enterprise software. Recently, a groundbreaking invention titled “Query system using multiple AI agents for text-to-SQL and text-to-Python” (U.S. Patent 12645672) was introduced by Zscaler, Inc. This innovative framework allows users to query databases using natural language, automatically generating the corresponding SQL or Python code without requiring deep technical expertise.
This remarkable development has rightfully earned the title of Patent of the Month for the AI, software, crypto, and cloud industry for July 2026. What makes this patent so innovative is its multi-agent approach to artificial intelligence. Instead of relying on a single large language model that might hallucinate or produce inaccurate code, the system deploys a collaborative team of specialized AI agents. This includes a research analyst to interpret the query, a search data engineer to execute the search, a code developer to write the SQL or Python scripts, and evaluator agents to verify the accuracy. This autonomous yet collaborative ecosystem ensures unprecedented precision, making it an essential tool for secure cloud environments and complex data sectors.
Transforming Enterprise Intelligence
The traditional barriers to data exploration often limit business intelligence capabilities, leaving non-technical staff dependent on specialized engineers. By implementing a semantic layer and leveraging multi-agent workflows, this invention democratizes data access. The AI agents work in a group chat environment managed by a chat manager, coordinating tasks seamlessly to fetch both general data and rich visual analytics through Python.
R&D Tax Credit Eligibility in the USA
For technology companies looking to implement or build upon the practical applications of this multi-agent framework, there are significant opportunities to claim the Research and Development (R&D) tax credit in the USA. Under IRC Section 41, businesses can claim credits for expenses incurred while resolving technological uncertainties through a process of experimentation. Developing proprietary multi-agent systems, integrating these advanced AI architectures into legacy database environments, or engineering hybrid retrieval models all require rigorous iterative testing. Documenting this development process, specifically how software engineers test and refine AI interactions to eliminate query errors, satisfies the four-part test for the R&D tax credit, offering a substantial financial incentive for continued innovation in the AI sector.
