Discover potential issues ​from documents using AI

We developed an AI Pre-approval and Cross-checking system to automate the RISC form verification process. The system identifies quality issues captured in the RISC forms by employing sentiment analysis and recommends inspection result (outcome), namely “Endorsed”, “Failed”, and “Cancelled”. The system counter-checks if a RISC form has been endorsed incorrectly based on the content of comment fields and provides reason for the prediction. Users can edit the result to retrain the model for continuous enhancement of accuracy. Besides, a word cloud is developed to provide users with a quick overview of their text data. The word cloud does not only consist of one word, but also phrases which frequently occur in the text data. This helps to provide a clearer picture of the overall text data and make it easier to identify patterns or trends in the data. Moreover, the system provides a natural language search function which allows users to express their queries in a more natural and intuitive way.