Banya AI Accelerate Market Expansion Across Financial and Public Institutions With CodePilot, An AI Coding Agent Specialized for Internal Networks

[SWTV 이지한 기자] Utilizing CodePilot, an AI coding product optimized for internal networks, Banya AI has launched a full-scale entry into the financial and public sectors.

Banya AI is a deep tech startup founded last year. As many companies and public institutions struggle to adopt generative AI owing to security and network separation issues, Banya AI provides AI services tailored for organizations restricted from using external, frontier LLMs.

Through the Seoul Startup Hub Gongdeok Open Innovation Program, the company collaborated with the Nonghyup Bank Card Business Division to conduct a Proof of Concept for natural language to SQL (NL2SQL) and data generation. The collaboration verified the feasibility of applying ‘CodePilot’ within a real-world financial environment and received an award for the Open Innovation project.

‘CodePilot’ is an on-premise AI coding product designed to support development by incorporating enterprise-specific environments, legacy systems, and team-level rules and policies.

CodePilot is designed to enhance the development productivity and consistency by implementing project and team-level policies while offering scalable configurations ranging from 27B to 120B parameter models. Additionally, hybrid deployment is also available for clients with established cloud-based LLM infrastructure.

Banya AI is accelerating the commercialization and expansion of CodePilot building on its experience.

Recently, Banya AI entered into an exclusive distributorship agreement with ITCEN CLOIT for CodePilot. It will leverage its AI coding technology to jointly develop and commercialize the AI coder module within ITCEN CLOIT’s multi AI agent management platform, AgentGo to accelerate entry into the financial, public, and enterprise markets. It is also strengthening its business expansion with a focus on implementation and integration tailored to actual customer environments.

The company is also simultaneously conducting research on its own proprietary AI coding models. Building upon its participation in the Qualcomm AI Innovator Program, it has accumulated research experience in snapdragon chipset-based inference and model optimization. It aims to strengthen its capabilities in developing lightweight and optimized AI models tailored for enterprise environments.

Hwang Byung-dae, the CEO of Banya AI, noted, “The company focuses on delivering practical AI coding agents tailored to specific organizational policies and systems, moving beyond the capabilities of a simple code generator. Building on the PoCs conducted in the financial sector and partnerships, we will continue to scale the market adoption.”

In fact, Banya AI aims to build references in the financial and public sectors, enhancing AI coding agents that ensure stable and secure management within internal enterprise networks.