Transforming Government With Data

Analytics and AI bring powerful insights to agencies and empower residents with information.

Constituents deserve data-backed evidence that government programs are paying off. That’s the principle behind an initiative that’s helping the state of Tennessee improve outcomes in areas like child services, higher education, corrections and substance abuse.

Tennessee’sOffice of Evidence and Impact (OEI) is working with Google Cloud to implement TN DATA (Data Analytics for Transparency and Accountability), an enterprise data analytics platform for evidence-based budgeting and data-driven policy insights. The office also offers a suite of programs that let residents tap into databases and access information amassed in agencies across the state.

Efforts underway at OEI reflect the growing appeal of using analytics and artificial intelligence (AI) to democratize the power of data. Tennessee is one of many states across the U.S. that are driving a cultural shift that’s transforming how government agencies engage and serve the public.

The impact of data technologies took center stage during “The Rise of Analytics,” a session at Government Technology’sTennessee Digital Government Summit that explored how advances in AI and machine learning (ML) are shaping the next phase of data-driven government.

Lauren Collett, an account executive with Google Cloud, urged public-sector IT leaders to use intelligent algorithms to move beyond collecting and curating data, and concentrate on generating useful insights and analysis.

“Data's kind of in my DNA,” Collett told attendees at the May summit. “As a taxpayer and native Tennessean, I like data to be used well for allocating our resources and driving policy.”

LOCATION DATA: A POWERFUL USE CASE

The session also explored how advanced analytics tools help state agencies use data from geographic information systems (GIS) to serve residents’ needs, which can vary widely from Memphis to Nashville to Knoxville (and everywhere in between).

“Tennessee is a very diverse state in terms of industries and population needs — almost three states in one,” Collett said. “Education and workforce programs in agricultural areas may be very different than in automotive and manufacturing locales.”

GIS data helps power some of the most attractive use cases for advanced analytics, AI and ML. Geospatial imagery, for instance, can help optimize broadband and infrastructure expansion by revealing the precise locations of structures on properties like farms and ranches — key information for delivering last-mile connectivity. GIS maps can also add layers of demographic data such as average income and education levels to help jurisdictions understand issues, target resources and strengthen economic development projections.

As the technology evolves, video data intelligence from cameras posted along streets and highways may help agencies improve traffic flow, Collett added.

OVERCOMING DATA OBSTACLES

The COVID-19 pandemic proved that agencies could avoid turf battles over data and work together to solve serious problems. One pandemic-related project cited in the Tennessee session used AI to predict the spread of COVID-19 in populations around the world for up to three months with 93 percent accuracy.

Advanced analytics technologies can also provide insights that help agencies improve constituent services. In addition, these tools can enable individuals and groups to access and examine data themselves, gaining power previously reserved for elected officials and agency leadership.

While these innovations are encouraging, agencies must also address potential data bias and governance issues as they broaden their use of analytics.

“I'm very passionate about data bias,” Collett said. Agencies need to consider the data they won’t collect in addition to the data they will amass, she said. “I know Tennessee is going to be very prudent in what they do with the insights from TN DATA.”

Data governance issues can present another hurdle. For example, compliance standards often forbid, or narrowly limit, moving data from one location to another. These rules make it difficult to run AI operations that require data to be transmitted over a network.

An approach called federated AI can help agencies overcome this obstacle. AI algorithms essentially travel to the data, collect results and transmit only the insights back to the application that requested the information.

The data never moves, so it can’t be exposed to security and privacy threats in transit. Moreover, insights drawn from multiple data sources can be scanned for anomalies and inconsistencies to help flag bias and improve accuracy.

EXCITEMENT OVER AI AND ANALYTICS IS BUILDING

The democratization of data is enabling a cultural shift. “In Tennessee, we’re excited to see this data culture forming,” Collett said.

The pervasive availability of cloud tools and multi-cloud architectures is dramatically expanding the potential of analytics and AI. Deploying servers and other hardware to enable data-intensive projects used to be extremely difficult. Now, analytics capabilities delivered through the cloud reduce the barriers to entry and let agencies realize benefits faster. Agencies can integrate modern cloud-based tools with existing systems to accelerate modernization and achieve quick results.

“Maybe we need to work through some nuances, but it's not a long effort to go buy infrastructure anymore since the technology is securely hosted and scales on demand,” Collett said.

In Tennessee, data and analytics are improving constituent services and program outcomes, while giving hope to distressed communities.

“Data is going to be required to address these issues,” Collett said. “It's going to take everyone's participation. If we focus on the outcomes, we can make some meaningful impacts.”

Visit Google Cloud’s State and Local Government Solutions site for more ideas and information.