Security · October 10, 2026
"Personal data leakage prevention"... BiMatrix targets large enterprises and financial sector with closed network AI
On October 8, 2026, BiMatrix showcased its on-premises 'Text-to-SQL' solution at the AI Festa held in Gangnam, Seoul, amid growing concerns over information leakage in large enterprises and the financial sector. This solution operates securely within corporate internal networks, maximizing security with its closed network AI technology.
BiMatrix, a company specializing in AI and BI solutions, utilizes generative AI and low-code technology to support data analysis and business system development for enterprises. The 'Text-to-SQL' solution converts user inquiries in natural language into SQL queries, allowing real-time access to data from internal databases and visualizing it on UI dashboards.
A key feature of this solution is its ability to completely eliminate the hallucination phenomenon commonly associated with generative AI. Instead of processing external data and creating fabricated results, it generates only accurate SQL statements that retrieve real data directly from the database.
User convenience has also been significantly enhanced. Previously, complex prompts were required to obtain desired results, but now, with prior knowledge of database schema information, even simple queries like "Tell me the status of complaints" can generate comprehensive dashboards automatically. This advancement drastically reduces the time needed for users to request data extraction from IT departments, which previously took several days.
BiMatrix emphasizes the 'closed network security' of its solution, which operates solely within corporate intranet environments without external online network connections. This has garnered strong interest from public institutions, financial organizations, and large manufacturing companies hesitant to adopt cloud AI due to recent data leakage issues.
Furthermore, the solution offers extensibility, allowing integration with existing MLOps environments or Python code for advanced computations, such as predicting signs of equipment anomalies. A representative from BiMatrix stated, "There is a growing demand for on-premises AI in the security-sensitive sectors of finance, public service, and large manufacturing, aiming to provide an optimal environment for immediate access to corporate data without security concerns or hallucinations."