Moody’s is a 100+ year old company that helps businesses and financial leaders better evaluate risk. From cyber to environmental to credit risk, Moody’s provides data and analytics for companies and individuals to make better decisions.
Moody's AI
was an AI chat-based platform where Moody’s employees could log in to analyze systems of risk data, compare document themes and manage files, search databases, and more, all powered by OpenAI’s LLM.
Building Moody's AI would contribute to the business's OKR of adopting new technologies to position Moody's as a leader in AI in the finance industry.
Testing a concept
early on with daily active users showed improved comprehension, better usability, and positive emotional reactions to the new design. I collaborated with a UX researcher to create both moderated and unmoderated tests on the Great Question platform to test out with participants we recruited.
We learned that
the home page I designed wasn't adding much value to the experience; it was in fact hindering people's desire to interact immediately with the chat because it required them to click once to navigate to the chat page. Thus, we eliminated the home page and focused on improving the CTAs for other areas of the workflow, like creating datasets and sharing them.
The new design made it easier
to create datasets and upload documents because we made the setup for access permissions a separate workflow. Doing so decreased cognitive load for non-admin users who wanted to upload documents first and share them second. A separate share and create document flow also mirrors how sharing works with files in Google Drive, Dropbox, and Figma. Employees could share documents via AD groups or by individual email.
14,000+ employees
saw the release of Moody's AI v2 in March 2025, which positioned Moody's even more as a leading innovator in AI in finance. Harvard Business Review subsequently published an article on Moody's AI.
Thank you to Tudor Bay, Denis Somar, Patricia Yu