Executive Summary
Mustard Golf uses AI and computer vision to analyse golf swings in real time and provide personalized video coaching to help users improve. While Mustard offered great swing analysis, they couldn’t scale personalized article recommendations from Golf Digest to match users’ swing issues. They partnered with Elastiq and Google Cloud to deploy a Gen-AI-powered content recommendation system that receives swing insights and delivers the most relevant Golf Digest articles to each user. This enhanced Mustard’s training experience, boosting engagement and retention.
Client Profile
Mustard delivers personalized training across multiple sports. Its product, Mustard Golf, helps amateur and professional golfers improve through real-time swing analysis and coaching. In partnership with Golf Digest, Mustard Golf has access to 3,000+ premium articles spanning over 75 years of golf training knowledge, enriching the user experience.
Business Problem
Mustard Golf relied heavily on video-based coaching powered by their swing analysis technology. While this provided personalized feedback, users lacked supplemental learning resources tailored to their specific areas of improvement.
Though Mustard had access to Golf Digest articles, manually curating relevant content per user swing profile was not scalable.
Mustard needed an AI-driven solution that could surface the right article at the right time, based not only on user engagement but also on the performance insights generated from their swing analysis.
Elastiq Solution
Elastiq built a scalable, serverless Gen-AI based Content Recommendation Engine, powered by Gemini on Google Cloud Platform (GCP), tailored to Mustard’s swing analytics pipeline.
Indexing the Data
- Indexed 3,000+ articles across 12 golf training variables (e.g., backswing, posture)
- Generated vector embeddings to capture the semantic meaning of training content and articles
- Used vector matching to recommend articles based on categories
Real-time Recommendations
- When swing analysis flagged a weakness (e.g., backswing issues), AI identified relevant training topics
- The system retrieved and recommended the best articles, and generated a summarized recommendation using Gemini. This complemented the video coaching content
- Recommendations were posted to the users in the Mustard App
Results
The transformation delivered measurable improvements:
- Higher Engagement: Personalized article recommendations enhanced learning alongside video drills
- Better Retention: A more comprehensive training experience keeps users active longer
- Scalability & Security: GCP’s managed services enabled scaling content delivery to thousands of users with high reliability
- Faster Innovation: Gemini models enabled deep semantic understanding and flexible integration with swing data
“Mustard users are now getting meaningful articles, personalized for their swing, directly into their Mustard plan. Users have already reached out to let us know how much they appreciate the articles.”
Conclusion
By leveraging AI for content curation, Mustard delivers personalized, scalable training that keeps golfers engaged and improving their game. This case study demonstrates AI’s power in transforming training platforms, driving engagement, efficiency, and impact across industries like sports, education, and enterprise learning.