‘AI skills factory’ created by Thomson Reuters for non-engineers to build expertise

Thomson Reuters’ recently released generative artificial intelligence (AI) platform features an “AI skills factory” that allows subject-matter experts who are not engineers or software developers to safely experiment with building new AI tools to address business needs.

“When we look across our customer bases in legal professionals, tax professionals, corporations or Reuters News, we see a huge amount of opportunity to apply generative AI,” Shawn Malhotra, head of engineering at Thomson Reuters, told FOX Business. “The problems that generative AI is good at solving are the same problems our customers are wrestling with today, so there’s a ton of opportunity.”

“However, we also know that we’ve got to move quickly, and we only have so much talent in the world who really understands generative AI and how to leverage it in solutions. So we need to move faster with limited resources, so how do we go off and do that? And that’s where the platform really comes in,” he explained.

To help broaden the pool of workers who can help find innovative AI-driven solutions using the platform, Thomson Reuters’ generative AI platform aims to make it easy for those who are not software engineers or developers to create those AI building blocks.

THOMSON REUTERS LAUNCHES GENERATIVE AI TOOLS FOR LEGAL RESEARCH

“What we’ve done over the last several months is we’ve added new building blocks, ones that help you build generative AI skills,” Malhotra explained. “Ones that basically abstract some of the complexities and nuance of generative AI so that any developer, and in some cases even subject-matter experts who aren’t coders, can start to assemble and experiment with generative AI.”

“That does two things; it lets us build applications much faster. It also allows more of our employee base to actually get involved with creating these solutions because sometimes it’s the deep domain expert who has the best idea who’s going to build the best solution,” he added. “So the platform enables them to build what we call an AI skill to figure out whether or not it’s adding customer value, so you increase the number of people who can participate in the innovation, and you increase the pace at which we can turn that innovation into real customer value.” 

WHAT IS ARTIFICIAL INTELLIGENCE (AI)?

Malhotra said that users of the platform who are looking to build an AI skill can choose one of Thomson Reuters’ proprietary content sets along with a language model and then experiment to see if it works as a solution to the problem they were trying to solve. 

He added that the Thomson Reuters platform’s AI skills-building platform is sandboxed to prevent mistakes like putting proprietary content in an externally accessible environment, while it also has built-in tools for data governance and detecting bias in AI models.

“We let them do all of this in an intuitive, low code, no code environment. They can then see the results of that, and because they’re the subject-matter expert, they can say ‘huh, that really looks like it would be great for a customer, that would solve a lawyer’s problem, a tax professional’s problem, a compliance officer’s problem.”

Once the user gets to that stage, they can engage the engineering team to turn those building blocks into a solution that the company can deliver to customers, Malhotra explained.

“Lots of folks are trying to deliver generative AI solutions, and what we’ve learned as we’ve now delivered some to market is that the secret sauce behind these solutions is really having great technology that you can build on quickly,” he added. “That’s where the Thomson Reuters generative AI platform comes in. But you also need the domain expertise and the content – those are truly differentiators in the solutions that are coming to market right now.”