The City of Boston is undertaking a multi-year effort to modernize its permitting system. The former process required residents to search for the correct permit application by sifting through administrative categories such as structural, electrical and plumbing. The new interface is designed with specific and common applications in mind, like replacing a boiler or building a deck. The city used a combination of Natural Language Processing (NLP) and a Large Language Model (LLM) to better understand historical patterns in permitting records. Staff applied those insights to create a new permitting portal that helps residents apply for the right permit with less confusion and friction.