Chris built Amy, an AI-powered calorie tracking app, reaching $1,500/month in revenue within two months by optimizing AI costs and improving user retention through innovative features and UX enhancements. By implementing cheaper AI models, caching, and unique tools like menu scanning, he significantly increased profit margins and user engagement, setting a strong foundation for future growth.
Chris, the creator of the AI-powered calorie tracking app Amy, shares his journey of building and growing the app to $1,500 in monthly recurring revenue within just two months. Amy stands out for its frictionless user experience, allowing users to simply type what they ate and instantly see calorie calculations, similar to Apple Notes. Despite the early financial success, Chris faced significant challenges with high AI costs and low user retention, which initially made the business model unsustainable.
To address the high operating costs, Chris implemented two major changes. First, he introduced a cheaper AI model, Gemini 2.5 Flashlight, for simple food portion edits, reserving the more expensive Perplexity Sonar model for complex queries. Second, he added caching using Supabase, so repeated food entries would pull from a database instead of triggering costly AI calls. These optimizations reduced his AI bill from $700 to $221 per month, even as the user base grew, with caching alone accounting for a 70% reduction in costs.
With improved profit margins, Chris felt confident enough to add innovative features, such as menu scanning. This feature allows users to take a photo of a restaurant menu, have the app extract dish information, and quickly log items with calorie counts. Powered by Gemini 2.5 Flashlight, this feature is unique among calorie tracking apps and creates a “magic moment” for users, further differentiating Amy from competitors.
Chris also focused on improving user retention, which he tracks closely using PostHog analytics. Initially, only 3% of users were still active after one week, but through continuous improvements, he raised this to 10%. Key changes included multiple small UX enhancements, explicit editing options for food photos, proper weight tracking with progress photos, improved data export and Apple Health integration, and smarter, less intrusive notifications. These updates aimed to reduce friction, increase user investment in the app, and build trust.
Overall, after two months, Amy has achieved a healthy profit margin of about 85% on AI costs and continues to grow in both revenue and user engagement. While not yet a massive success, Chris is optimistic about the app’s future, driven by ongoing learning and a clear roadmap for further improvements. He encourages viewers to follow his journey on social media and subscribe for more content on building productivity apps.