The growth in data is outpacing our ability to understand it. There are plenty of tools that ought to help, but they are beyond the reach of general users.
To glean anything meaningful from data, you need specialists with SQL and Python expertise to produce even simple two-dimensional reports, such as the number of disease infections over time or total Medicaid applications by region.
Even then, users tend to get bogged down. A recent report by the Johns Hopkins Program in Government Analytics found that data professionals spend almost a quarter of their time gathering data and only 8% analyzing it.
Analysts say that’s unlikely to change anytime soon. Research firm IDC expects that the amount of data created annually will reach 175 zettabytes in 2025.
The government is among the world’s biggest data generators and consumers, but without a way to make the data useful, it’s just bits and bytes.
“As the gap between the amount of data available and the number of trained experts who can analyze it grows, federal agencies must empower employees at all levels to use their data to be their own data specialists to get timely insights,” said Monica McEwen, Vice President of Public Sector at ThoughtSpot.
Artificial intelligence (AI) is seen as essential to getting more value out of data. But there’s a skills gap.
Respondents to the Johns Hopkins survey said that staffing is the biggest hurdle to using AI. Although a quarter of them said AI holds the most promise for improving government, less than 4% work with the technology now.
In 2019, the American AI Initiative directed the Select Committee on AI to provide education and workforce recommendations for developing an AI-ready workforce. But government workers, who are already pressed for time, don’t have hours to sit through training – and agencies, already pressed for funding, don’t have the money to provide it.
The answer is a search- and AI-driven analytics platform, which enables everyone to use data for actionable decision-making.
A modern analytics platform will enable non-technical users to analyze data, with Google-like search, powered by AI and machine learning. This will let government workers of all technical skills search for data and automatically get answers in the form of visualizations. Imagine being able to perform a Google-like search such as “What’s the number of positive COVID-19 cases in my state?” and watch as the answer pops up in graphical form.
Indeed, using AI could free up 30% of the government workforce to do more mission-critical work, according to a report from the Deloitte Center for Government Insights. That work is beginning to happen: A report by the Administrative Conference of the United States found that 45% of federal agencies have experimented with AI.
Still, public-sector IT leaders estimate that 56% of their data is “dark,” meaning it’s unknown or untapped, a survey by Splunk found.
Business intelligence is not a new capability, but traditionally it’s been designed for technical data users. ThoughtSpot’s solution, however, democratizes data through AI. Anyone at the agency can analyze billions of rows of data across the organization in seconds by querying the company’s relational database, just as they would use a search engine to get information. ThoughtSpot’s AI engine, SpotIQ, returns insights as you search – even answering questions you didn’t know to ask.
The key to this self-service analytics is the combination of a simplistic user interface and AI to better enable the user experience. ThoughtSpot’s Relational Search and AI Engine work to identify areas of the most importance and also the biggest outliers in the data, which could tell a story of their own. The result is significantly faster – and fact-based – decision-making.