TransForm: Generative AI for Filling Forms

Manually filling in forms is an activity that is pervasive in both Government and commercial activity. Most forms must be filled manually, which is time consuming and prone to error. Technology that assists form filling would save significant time, effort, and money, while also improving the quality of information in the forms and the decisions that are made based on them.As a specific example, maintenance of ships and their equipment is essential for Navy mission readiness. Tracking maintenance issues is accomplished using 2-Kilo maintenance action forms. Information entered on 2-Kilo forms flows downstream to systems that order parts and assess readiness. These 2-Kilo forms contain many fields, some of which are free-form text, and are filled out by Sailors who have limited time for paperwork. Any shortcuts or omissions can lead to incomplete or incorrect information flowing to the downstream processing. The Navy would benefit from technology that reduces the time required to enter data into 2-Kilo forms accurately and completely. Charles River Analytics offers TransForm, a system that uses large language models to assist form filling, with 2-Kilo forms as the first test case. TransForm enables fast form filling while preventing omissions, ambiguities, and redundancies that may disrupt downstream processing.