Jorge Sandoval@jorgesandev

US hospitals leave hundreds of billions in recoverable revenue on the table every year because writing insurance appeal letters is too slow and expensive to scale. DenialDefender is the autonomous appeals layer designed for hospital revenue cycle management (RCM) teams and third-party RCM services. A user uploads a denial letter, and our multimodal system ingests the document, the patient's full chart, the payer's medical policy, clinical literature, and past successful appeals. Within 60 to 90 seconds, it drafts a complete, submission-ready appeal packet, including a medical necessity letter, code corrections, citations, and an overturn confidence score. Because this is a massive memory-bandwidth-bound workload requiring context windows up to 64K tokens for the demo (scaling to 200K+ in production), we rely on the AMD Instinct MI300X. We run both a primary reasoning model (Qwen3-32B) and a vision model (Qwen2.5-VL-7B) co-resident on a single GPU in FP16. This drives our end-to-end compute cost down to just $0.03 per appeal, enabling our core contingency business model: we don't charge if we don't recover.

Grab your ideas, spin up compute, and start building what’s next! Start with $100 in AMD Developer Cloud credits and begin building right away. Get access to hands-on tutorials, workshops, and curated learning resources to help you build faster and go further. To unlock credits and join the hackathon, sign up for the AMD AI Developer Program. 🤝 Go solo or team up — ship something real. 💰 $21,500+ total prize pool & 1 × AMD Radeon™ AI PRO R9700 GPU up for grabs. 📍 On-site Venue (May 9-10): MindsDB SF AI Collective 3154 17th St, San Francisco, California, USA 📍On-site participation is by invitation only. Travel and accommodation expenses will not be covered. 🧑‍💻 Developers, builders, and innovators - bring your boldest ideas and let’s turn them into real AI systems on AMD.