For 15 years, infoSentience has created AI-powered data-driven articles and reports in natural language at scale.
Between its powerful customization tools for producing personal narratives for each end user and built-in human touchpoints that give stakeholders more control, infoSentience instantly unlocks the power of one thousand analysts working at lightspeed—surfacing compelling information that might otherwise slip through the cracks.
infoSentience offers users powerful AI with expertly engineered guardrails to ensure accuracy and high-quality output. Unlike popular large language models, infoSentience is built upon a framework of non-probabilistic AI—one that ensures both depth and accuracy of reports without hallucinations. In addition to narrative articles, infoSentience features formatting tools that allow users to automatically add visuals like charts and graphs to reports.
Founder and CEO Steve Wasick was in his third year of law school at Northwestern and barely paying attention to his fantasy football league when he was struck with inspiration: “I liked following sports and reading updates, but I didn’t have a way to follow my fantasy league like real sports teams. I thought about how it would be cool if there was a way to have personalized coverage for your fantasy team, so I got to work.”
Nine months of building and one small friends-and-family fundraising round later, Steve had built his first version of infoSentience (then known as Fantasy Journalist), hired his first employee, and landed his first client: CBS Sports. infoSentience has delivered more than 200 million unique automated fantasy sports reports and game previews and recaps for CBS Sports and its millions of fantasy sports players.
infoSentience has also applied its proprietary technology to generate millions of articles for MaxPreps, the leader in coverage of American high school sports, providing coverage for countless events – and athletes – that wouldn’t have been highlighted otherwise.
“I wanted to move beyond the mad libs approach others had taken and build out intelligence, instead of simple logic,” explains Wasick. “More than just following rules like ‘if this, then that’—I wanted the model to understand how it got from Point A to Point B. I realized it would have a broader application once I achieved this.”
Wasick put his head down and spent eight years perfecting infoSentience’s AI—starting from scratch seven times before finishing the model in 2021. The final product: a tech stack that withstands the test of time—still supporting fully bespoke, configurable, and custom-built solutions for high-profile clients today. For fantasy sports leagues, infoSentience can generate millions of reports customized for each end user and write articles as if each team existed and played that week, calling on player stats and historical data – both real-world and inside each fantasy league – to deliver data-driven bragging rights.
In 2017, infoSentience began producing reports on the NBA and NFL for CBS as well as sports betting content. The next year, the company started working with the Chicago Mercantile Exchange, turning raw data into narrative reports on commodity markets for the largest commodities trading firm in the world.
“When we began this journey, infoSentience was one of just a few fish in a very small AI pond. Today, the masses have access to large language models, but with millions of parameters, the likelihood of hallucinations and misinformation is too risky for many organizations to use them. infoSentience provides a viable option for businesses that want a powerful AI model for automated reports that they don’t have to worry about,” explains Wasick. “From cross-checking the accuracy of LLMs to generating millions of reports at scale for training other models, non-probabilistic AI like infoSentience will play a critical complementary role in building the future of AI.”
As the industry discovers the limitations of LLMs and other trendy AI approaches, infoSentience will remain at the forefront of innovation—building powerful models that prioritize accuracy, depth, and efficiency.
