English 
AI inside Inc. has announced the development of ‘Critic Intelligence,’ a proprietary error-detecting AI, and its integration into the company’s DX Suite, an AI agent designed to automate data entry processes. The new technology tackles a key problem in document processing. It greatly boosts accuracy and efficiency in finding errors.
Traditionally, AI-OCR digitized forms needed a lot of human checks. This often meant several reviewers to ensure accuracy in finance, insurance, and BPO. This manual process was time-consuming and prone to oversight. Critic Intelligence spots items likely to have errors. This helps human reviewers focus on just the questionable entries.
Testing has shown the system can detect errors with an average accuracy of 99.86%, far surpassing the 60–80% accuracy rate of manual checks. In addition, processing times were reduced by roughly 80%, enabling organizations to achieve both higher quality and faster turnaround in data entry workflows.
こちらもお読みください： Tokio Marine Partners with OpenAI to Deploy AI Agents
The launch of Critic Intelligence marks a major step toward 内部のAI’s vision of AI as a collaborative partner rather than just a tool, setting the stage for more advanced applications in the era of artificial general intelligence.
日本語