The AI agent technology corporation 'Smart Mind' announced on the 30th that it has launched its self-developed natural language query and response solution 'Qurify'.
This product is a solution designed so that regular employees in non-IT departments can query databases, Excel, and CSV files in natural language and receive accurate responses.
Qurify is characterized by its ability to maintain the highest accuracy level without the need for frequent updates of table information or model fine-tuning that existing commercial and open-source solutions required. Most existing solutions needed to build their virtual large language models (VLLM) to generate accurate queries or required ongoing model updates, which incurred high expenses and technical burdens for implementation and maintenance.
Smart Mind's Qurify fundamentally addresses these issues. Customers only need to provide the meta-information of the tables, and the dedicated server automatically generates SQL (Structured Query Language) for natural language queries based on this, allowing it to handle most complex queries that were not executed by existing solutions.
Additionally, Qurify is equipped with a large language model (LLM) based on a Chain of Thought reasoning that Smart Mind has tuned in-house, allowing it to automatically correct and execute queries generated incorrectly.
Moreover, by assessing the appropriateness of the query execution results and automatically improving them, customers can continuously receive high-accuracy services without the need for repetitive table information updates or model retraining.
Lee Sang-soo, CEO of Smart Mind, noted, 'Qurify eliminates the burden of repetitive updates and fine-tuning that existing solutions faced and ensures accurate responses even with queries in everyday language,' and added, 'It can reduce the waiting time for non-IT personnel to request data from the IT department by more than 200 times, cutting expenses and time while dramatically enhancing productivity.'