DRF AI: Untangling and visualizing complex horse racing information
The Daily Racing Form has been North America's primary source for thoroughbred racing statistics, past performances, and news since 1894 with currently over 3M+ visitors per month. At Haptiq, I led the design of DRF AI, partnering with an in-house AI engineering team.
I translated dense racing data into intuitive AI prompts and clear data visualizations, making complex past performances easy to explore and act on. I also built a complete design system for the product, cutting design-to-dev turnaround by 25%.
Client
Daily Racing Form
Year
2026
Scope
B2C· AI Feature · Design Lead
Summary: Horse racing data is powerful, but overwhelming.
Daily Racing Form (DRF) has been the go-to source for statistics, past performances, and expert analysis for bettors, handicappers, and racing professionals, yet newer users often struggled with the terminology, with comparing horses, and with feeling confident in their picks.
DRF AI removes that barrier by letting users ask questions in plain language and get clear, contextual insights in return.
Define Problems: Outdated Information Structure & Inconsistency
Users were often overwhelmed by the complexity of horse racing data presented as dense text, and newer users struggled most. Behind the scenes, the lack of a defined design system slowed design-to-engineering handoff, increasing both cost and effort and limiting the team's ability to move quickly.
Challenge & Goals: IA & Design System
I first recognized the urgency of restructuring the information architecture and visualizing the racing numbers. After more than a month of discussions with stakeholders and interviews with users, I designed the main display for the information cards.
This became the foundation for all future development of DRF AI's visual display.
Prompt-First UX
I built the first prompt framework by defining the role, situation, and task for each scenario, working with UX writers to make sure the prompts were accurate.
Prompt Augmentation & Proactive Guidance
A general chat function paired with prompt suggestions gave new users reassurance and helped guide them to get started. Actionable items generated in the responses increased click-through rate by over 33% in our initial run compared to the previous text only layout.
Shift to Intent-Based Outcome Specification
Shifted to intent-based prompting, which lets the team tailor responses to each user's intent. This produced more relevant suggestions and, in turn, more actionable items.
Impact & Numbers
DRF AI is now live to the site's 3M+ monthly visitors, driving a 5% click-through rate on revenue-generating items and 6% on related ads and news. Ads sections are currently generating over 1M in revenue per month
AI Integration in an Extended Product Line: Newbie
Newbie is DRF's extended product line designed specifically for new users. Because the design system was already in place, DRF AI integrated into Newbie smoothly and consistently.
