SisunAI Launches sLLM-Based Coding Automation AI Solution Targeting Public and Financial Institutions

Vision AI specialist SisunAI has launched a new enterprise on-premise (closed) AI Code Assistant solution, "IntraGenX 1.0," which is based on a small large language model (sLLM) specialized for coding automation.

SisunAI's "IntraGenX" is a solution that supports and extends developers’ coding work through natural language prompts. It is a product jointly developed in collaboration with its subsidiary Daebo DX. The solution features collaborative AI pair programming and prompt-driven development capabilities, delivering high productivity in practical development tasks such as logic design, code review, bug detection, and correction.

The core function of IntraGenX is the automatic generation of models and code directly from development design intent. When data models (ERD), design files, or other outputs already exist, these can be integrated to generate more precise code (both frontend and backend) automatically. For frontend development, it supports global standards such as React and Vue, as well as major domestic UI platform environments, enabling optimized results immediately.

Traditional frontend development methods required multiple steps—including screen design, HTML publishing, and JavaScript (JS) coding according to standard environments—resulting in significant time and workforce consumption. IntraGenX from SisunAI dramatically reduces the complexity and resources required by existing methods, enhancing productivity; automatically generates code in batches based on design documents, minimizing human error; and enables consistent, high-quality code even in environments with extensive development teams. As a result, it is expected to greatly contribute to productivity innovation and quality management, especially in next-generation development work at public and financial institutions where hundreds or thousands of personnel may be involved.

Additionally, this product is a closed solution supporting air-gapped environments, reducing security risks compared to open (cloud-based) LLMs. In the case of open LLMs, concerns have been raised about the leakage of sensitive internal information, prompting large enterprises to increasingly restrict and control employees from using such tools for work. IntraGenX adopts an on-premise approach, which enables a significantly higher level of security management, including blocking data leaks, ensuring data sovereignty, and regulatory compliance.

SisunAI has completed sLLM technology that, while streamlining and optimizing the LLM architecture with vast parameters for enterprise environments, enhances performance for special purposes (coding) through proprietary fine-tuning and AI training capabilities. As a result, IntraGenX delivers a high level of contextual understanding and precise code generation performance, even in relatively lightweight runtime environments.

Nam Woonsung, CEO of SisunAI, stated, "Despite the many benefits and expected effects brought by generative AI and LLMs, many institutions and companies are forced to forgo these advantages due to security concerns, which in turn exacerbates the productivity gap with other organizations that are less restricted. IntraGenX will be a tool that can significantly enhance productivity, efficiency, and accuracy in development work while ensuring robust security."

SisunAI plans to expand the specialized market for IntraGenX, focusing on enterprises and institutions requiring high levels of security, such as public and financial organizations, in collaboration with its AI transformation (AX) and system integration (SI) specialist subsidiary Daebo DX. As a strategic agent, Daebo DX will be responsible for discovering new clients, sales, on-site system implementation, and tailored technical support. SisunAI aims to rapidly establish IntraGenX in the LLM market and drive performance by leveraging Daebo DX's expertise and network.

Meanwhile, the LLM market is showing strong growth as it moves from the technology adoption phase to enterprise solutions (Enterprise LLMs). According to market research firm Precedence Research, the global LLM market is expected to reach approximately $10.57 billion (about KRW 15 trillion) this year, with a steep compound annual growth rate (CAGR) of 34.44%, and is projected to reach $149.89 billion (about KRW 219 trillion) by 2035. In particular, coding automation is being highlighted as a field where LLMs are driving real productivity innovation. Gartner forecasts that by 2028, 75% of enterprise developers will be using AI code assistants.

This content was produced with the assistance of AI translation services.