<p><b>Startup description</b></p><p>I'm a Louisville, KY native and this year, when the Kentucky Derby rolled around, I used AI to handicap some of the races for myself. Then, I turned that into a pick sheet that we sold to other people.</p><p><b>Key highlights:<b></b></b></p><ul><li>$13.8K in gross revenue in May 2025 - June 7th 2025</li><li>$8.6K profit</li><li>One time sales - but can turn into SaaS</li><li>430 unique customers</li><li>Pick sheet averages $5 - $25 per day / card</li><li>I've basically sunsetting the app and sale to existing email subscribers (740). We average anywhere from $30 - $80 a weekend selling the bigger stakes races</li><li>From June 26th - July 25th, we've done $343 in gross sales.</li></ul><p><b>Team</b></p><p>I mainly operate everything - setting up ads, building the pick sheets, and doing customer support.</p><p>As mentioned above, the only thing I do these days is run the cards via AI, and the email to sale the card.</p><p><b>Tech Stack</b></p><p>Program currently runs on a C# console app the builds the cards and outputs HTML. It pulls in scratch data and can modify track conditions based on weather - wet tracks handicap differently than fast (dry) tracks. Can swap or modify AI models here. We then distribute via email/SendOwl.</p><p><b>Marketing and growth </b></p><p>Facebook ads, and paid ads with media companies or networks. Again, we paused all marketing efforts after Belmont Stakes race first week of June.</p><p><b>Revenue and profit</b></p><p>$13.8K sales, $8.6K in profit in May 2025.</p><p><b>Return on investment</b></p><p>Our product returned 270 ROI in 'bettors' that followed the AI bet for Kentucky Oaks and Kentucky Derby. Returned 80% on Black-Eyed Susan Day, and lost about 50% on Preakness day.</p><ul><li>We picked/hit both exactas (bet type) for Preakness and Derby.</li></ul><p><b>Startup assets<b></b></b></p><ul><li>830 email subscribers</li><li>Facebook account and ads - ($0.04 - $0.08 per click)</li><li>Data contract - ability to resell data and use it with AI</li><li>Code to build the handicapped sheet.</li></ul><p><b>Risks</b></p><p>There are tons of big races, but we poured all marketing efforts into the triple crown races that are now over for this year. Focused on the largest to get a proof of concept.</p><p>Our data contract has monthly minimums if no sales are made.</p><p>AI competition entering the space is always a risk.</p><p><b>Summary</b></p><p>We proved the model, have buyers, and have folks already interested in pre-purchasing the Belmont. We ONLY sold the big races and did everything manually to prove out the concept. We proved out the concept, but decided we didn't want to be in the space for lots of reasons.</p><p>If I was to pursue this, I'd automate everything, and produce pick sheets for every track, get monthly subscribers and turn into a full business.</p></br>✅ $13,800 in Annual Revenue✅ 430 customers✅ Business model: One-time sales✅ Built with OpenAI, HTML, C#