City Tech Leaders Talk AI Implementation at TAGITM

Artificial intelligence was a hot topic at this year’s Texas Association of Governmental Information Technology Managers (TAGITM) conference in San Antonio, with multiple C-suite local government representatives sharing how their respective cities have implemented the technology.

On Wednesday, McKinney CIO Omar Rodriguez detailed his city’s progress while advising government leaders on how to safely and responsibly integrate AI.

According to Rodriguez, McKinney has made a concerted effort to embrace AI in multiple areas to maximize efficiency, such as training an AI chat assistant to help answer both employee and citizen questions either through chat or over the phone.

“I think this is kind of the potential, especially when we start thinking about the statistics,” said Rodriguez after demonstrating the chatbot’s capabilities. “Those folks that were answering the phone or answering those questions, they can now be applied somewhere else to do more important things.”

McKinney has also deployed AI to help with emergency response planning, particularly determining optimal locations for new fire stations.

Planning and zoning is the next area the city intends to enhance with AI, potentially automating permit applications by pre-filling forms and tagging applications completed by bots to streamline the approval process for planners.

On Thursday, city of Fort Worth CTO Kevin Gunn explained how government leaders can benefit from training generative AI on their city’s data, particularly though retrieval-augmented generation (RAG).

Gunn explained that cities can enhance AI models with specific facts without having to retrain the entire model through RAG, which stores information in text format, chunks it, tokenizes it and fine-tunes it for context. This allows an AI model to retrieve relevant information from a city’s data sources when generating responses, making its responses more accurate and easier to implement than fully retraining large language models.

As for how Fort Worth has implemented AI, Gunn detailed two pilot projects in partnership with IBM: a digital concierge chatbot and the streamlining of development application reviews.

According to Gunn, the city’s goal for the chatbot is to deflect 15 percent of calls to AI, and projects approximately $131,000 in savings during the next two years. The development application check is projected to save the city approximately $146,000 in the same period.

Fort Worth is also working with graduate students at the Neeley School of Business at Texas Christian University on using AI to review plan submissions for code compliance issues, which currently takes between one and 30 hours to complete without AI assistance depending on each plan’s complexity.

"With generative AI, there's a little bit of work to be done,” said Gunn. “But that doesn't mean that we can't look at these technologies, evaluate them for ourselves, find out where there might be some early application for an incorporated increase that we're doing because I do think there's a lot of potential for this technology."