Webcash to launch AI data integration for small business tax records

Connecting AI to private data

South Korean company Webcash announced that it will launch a new generative AI integration for its 'Kyungri Nara' business management platform in December. The update aims to help small and medium-sized enterprises (SMEs) access their corporate funds and tax invoice data through popular AI services like ChatGPT and Claude.

Currently, many businesses keep their sensitive financial records separate from general-purpose AI tools due to security concerns. Webcash plans to bridge this gap by using the Model Context Protocol (MCP), which is an open standard designed to connect AI applications to data held in external systems.

By using this standard, the platform allows the AI to pull the specific information it needs to answer a user's prompt. This replaces the current manual workflow where employees must download records into Excel files and then re-enter that data into an AI application to get any meaningful insights.

Automating financial queries

The primary goal of the Kyungri Nara MCP is to eliminate repetitive manual work for business owners and accountants. Once the integration is active, a user can simply ask the AI a question about their finances and receive a composed answer based on real-time data.

For instance, a user could ask the system to identify where spending increased compared to the previous month. The AI would then automatically compare current withdrawals against the prior month's records and list the specific counterparties where costs rose along with the size of those changes.

Beyond spending trends, the system will allow users to search for specific transaction records based on timeframes or amount conditions. It can also be used to look up purchase tax invoices by identifying the specific counterparty involved in a transaction.

Security and permission controls

Because financial data is highly sensitive, Webcash is implementing a structure where client companies maintain strict control over what information is shared. A connection to the AI is only established if a company administrator manually turns on the integration and provides explicit consent.

The system allows for granular permissions, meaning administrators can choose to expose only specific portions of their data. The available categories are divided into four distinct business units: funds, sales, purchases, and basic information. This ensures that only the necessary scope of data is accessible to the AI app at any given time.

The company also noted that these permissions can be revoked immediately if the administrator switches the integration off. This level of access control is expected to be a major factor in determining how quickly businesses adopt these types of AI tools.

Pilot program and future plans

Webcash will begin by selecting 10 companies from its existing Kyungri Nara client base to participate in a pilot group. Applications for this program can be submitted through notices within the service, and the selected companies will be able to use the MCP feature for one year.

As a benefit for participating in the pilot, these companies will receive a lifetime free subscription to the Kyungri Nara platform. They will also be asked to share their use cases to help the company refine the service before the full launch in December.

Looking further ahead, Webcash plans to introduce a dedicated AI agent called 'Kyungri 365' next year. While the current MCP focuses on connecting to general AI apps, this new agent will be specialized specifically for handling accounting and funds work within a dedicated environment.

Shifting the software landscape

Industry observers suggest that the rise of AI integration standards could change how people evaluate business software. The competitive focus for existing ERP and accounting platforms may shift from how well they store data to how easily that data can be used by AI models.

CEO Kang Won-joo noted that while many companies are already using generative AI, corporate funds and tax invoices have largely remained outside of these systems. He said they will continue to expand the scope in which management data can be used in AI.

This development specifically targets the needs of small businesses that have struggled to balance the benefits of AI with the need for data security. By providing a standardized way to connect private records to external models, Webcash hopes to make advanced financial analysis more accessible to a wider audience.