Local governments are using AI for a variety of applications, including the permitting process. The city of Sandy Springs has been using technology to better serve residents for years.
The midsize Georgia city is looking to fund the project with a Partnership for Inclusive Innovation grant, which it applied for with Georgia Tech in May. If the city is awarded the grant, which officials expect to learn in July, work will begin in September. If not, they will explore alternative ways to fund it, according to Keith McMellen, the city’s director of data strategy, analytics and AI integration.
The permitting process is a challenging one, McMellen said, noting the time it takes and the number of reviews, which was the catalyst for leveraging AI to review permits more efficiently.
The plan, he said, is to develop AI models that will review plans and permit documents to ensure they are accurate and complete — and, ultimately, quicken the review process.
The project will include the use of optical character recognition and image recognition technology to identify and extract key data points from documents, including those with handwritten information. Inconsistencies or missing information will be flagged for human review.
“It’s not to replace humans, it’s to just make things more efficient,” McMellen emphasized.
A successful AI implementation requires high-quality data, and for this endeavor, city officials are looking to train the models on existing data from previously submitted documents to learn where information is located and how to verify it.
As the mayor and council explore potential opportunities for AI technology to improve city services, the city has prioritized data management through initiatives including the creation of an internal Data Management Steering Committee. Through this cross-departmental initiative, the city can explore and address policy and priorities. Officials are also working to obtain a What Works City Certification, a recognition based on an assessment of a city’s data-driven decision-making.
Sandy Springs’ prioritization of data management ultimately led to the creation of McMellen’s position with the city, to provide expertise in this area.
“In all reality, data has become so much a part of our everyday lives,” McMellen said of the January creation of his role. “It was about formulating a strategy they needed to be able to address it at a citywide level.”
Data quality, he said, is a top priority to enable construction of high-quality, accurate AI systems.
“We don’t have a data problem,” he said, citing the widespread availability of data to which officials have access. “What we have a challenge with is data integration and data quality.”
He also emphasized that the city is not rushing to implement AI in every system, but rather, is starting with the permitting use case, for which high-quality data is available, and the impacts will be clear and measurable.
“We need to do it in a calculated, phased way so that we can prove there’s value,” he said, noting that by proving the city can use AI ethically and sustainably, future implementations can build on that.
To prove the value in the permitting initiative, the city plans to assess various metrics. The primary focus is to reduce the average number of reviews in the permitting process. Leaders also plan to assess the accuracy of AI models compared to humans, and the time spent to complete the process.
The approach McMellen said he recommends for cities, regardless of size, is to implement AI with a focus on solving a specific problem and providing a tangible, quantifiable benefit to people: “Don’t be afraid to get started.”
Sandy Springs is also launching a data and AI literacy program for staff, led by McMellen, with the potential to expand to other parts of the community like local schools.
