Smart Cities Connect 2026: Cities Test Small Before Scaling AI and Data Projects – StateTech Magazine

These 25 state and local IT leaders keep the lights on and the wheels of innovation turning in government.
The skyline of Trenton, New Jersey
Learn how AI is helping IT teams manage risk and improve resilience.

Jessica Balen is a freelance writer for the CDW family of magazines.
Cities do not need to start innovation projects with a large technology deployment, officials recently told the Smart Cities Connect Fall Conference and Expo.
In Trenton, N.J., Naman Sharma began by mapping how employees actually handled city-owned properties — and found that staff members were following different versions of the same process. That discovery shaped his approach to modernization: Understand the people and the process first, then decide what technology belongs in the solution.
“People, process, technology — and the order really matters,” said Sharma, an innovation lead and Bloomberg Harvard City Hall Fellow with the city of Trenton.
Sharma and Aathira Pillai, an innovation lead and Bloomberg Harvard City Hall Fellow with the city of New Bedford, Mass., discussed experimentation, analytics and artificial intelligence during the 2026 Smart Cities Connect conference. Both described projects in which cities tested smaller changes before committing to broader deployments.
Click the banner below for guidance on a data strategy for unifying stores and breaking down silos.
Sharma focused on Trenton’s stock of abandoned and city-owned properties. Nearly 10% of the city’s building stock was abandoned, he said, and roughly half of those properties were owned by the city.
The city’s process for disposing of those properties was inconsistent. Sharma interviewed employees involved in the work and found that each staff member had a different version of the process in mind.
The city created process maps, compared those workflows and brought employees together to develop a common approach. Sharma also developed user personas for both residents and city employees before the city tested a minimum viable product.
A technology vendor entered the project only after that work.
The testing exposed other problems before a full rollout, including unclear ownership, documentation gaps and an overworked real estate staff. Sharma said the city also considered how applicants might try to exploit weaknesses in the process.
“Learning while doing actually is the only design process that’s going to work,” he said.
Trenton eventually deployed a bilingual digital application system for the property program.
READ MORE: Here is a guide to stronger data governance for state and local agencies. 
Pillai found a different problem in New Bedford. The city had plenty of data, but department leaders were spending too much time collecting and organizing it.
She described performance management meetings built around large Excel files. The fire chief, for example, was manually working with incident coordinates to prepare information for the next meeting.
New Bedford shifted toward visual and geospatial analysis. Fire incidents could be mapped, clusters identified and discussions organized around specific problems rather than broad reporting.
Pillai said fires declined by more than 18% over two years as the city used that analysis to identify interventions.
The meetings changed as well. Departments began bringing forward the problems they wanted to solve instead of waiting for assignments from the mayor. Pillai also pushed for permanent data science positions so the work would continue after her fellowship.
LEARN MORE: Here is a guide to AI governance for state and local agencies.
Pillai used the same process-first approach when examining permitting.
Residents often arrived at New Bedford City Hall without knowing whether they needed a permit, which permit applied or which department handled it. Staff then spent time answering repeated questions and redirecting visitors.
The city created a decision tree system that lets residents identify the permits they need and move directly to the appropriate application.
Pillai said New Bedford also developed a citywide data strategy that became the city’s official three-year data plan. The city drafted an artificial intelligence policy, began training employees who had little experience with AI and started looking for routine work that could be automated.
For both fellows, experimentation meant finding problems while they were still small enough to fix.
In Trenton, that meant testing a property disposition process before bringing in a vendor. In New Bedford, it meant changing how officials used data before adding more technology to the system.
Artificial Intelligence
Q&A: New Jersey's Chief Innovation Officer Lifts the Curtain on Launching AI Projects
Security
How States Secure the Cloud Era with Identity-First Strategies
Visit Some Of Our Other Technology Websites:
Tap into practical IT advice from CDW experts
Copyright © 2026 CDW LLC 200 N. Milwaukee Avenue, Vernon Hills, IL 60061
Do Not Sell My Personal Information

source
This is a newsfeed from leading technology publications. No additional editorial review has been performed before posting.