From AI’s Role in Revolutionizing Technology with Perplexity’s Dmitry Shevelenko & FactSet’s Kate Stepp · · RiskReversalMedia
“Putting transcript assistant next to those sources that our users are used to using has made it really easy to dip your toe in and get in there, try it out — and also in a way that's kind of safer, with clear source facts and auditability in a way that our users feel comfortable doing so.”
On , Katherine Stepp, Executive VP & CTO at Factset Research Systems Inc, spoke about product safety during AI’s Role in Revolutionizing Technology with Perplexity’s Dmitry Shevelenko & FactSet’s Kate Stepp on RiskReversalMedia.
In a December 2024 webinar, Stepp discussed FactSet's approach to integrating generative AI into its financial data platform. She said that the company's history of integrating and collecting data provided a foundation for applying machine learning, and that generative AI has "revolutionized that efficiency factor." Stepp described how FactSet uses AI to understand user queries and then applies its existing, trusted query mechanisms to retrieve structured content, rather than relying solely on a pre-trained model for answers. She stated that this approach provides transparency and auditability, allowing users to trace information back to individual documents. Stepp also addressed user concerns about privacy and security, stating that FactSet does not train its internal systems on prompts from enterprise users. She noted that the company builds AI features in a compartmentalized way to allow for flexibility across different use cases. Stepp highlighted that AI-driven tools, such as a transcript assistant placed alongside familiar sources, have helped users adopt the technology in a "safer" manner, and that dynamic querying allows users to automate workflows like portfolio commentary, reducing tasks that previously took hours to minutes.