@croissantwithds885

SVAASA is a luxury Ayurvedic fragrance platform that replaces generic perfume shopping with a personalized, therapeutics-aware formulation and concierge experience — and puts the scoring math on the GPU. The core is a hybrid Formulation Engine: users pick health conditions, allergens, and scent preferences; the engine rule-filters out contraindicated and allergen-matched ingredients, then scores the rest by therapeutic relevance and preference fit. Scoring runs on a custom WGSL compute shader through WebGPU — each ingredient scored by its own GPU thread (64-thread workgroups), dispatched via Chrome's Vulkan backend onto AMD RDNA compute units, with a SIMD-friendly typed-array fallback when WebGPU is unavailable. A live benchmark page (/amd-compute) compares CPU vs. GPU timing across 64 to 16,384 ingredients. Scored ingredients are distributed across a base/heart/top note pyramid and ratio-adjusted by intensity and medium into a custom attar recipe. Around it sits the Aureān Concierge, a "zero-system-language" support channel: clients see only their message and a timestamp, never a ticket number, queue position, or SLA countdown. A client-side NLP triage classifier reads every message for intent (medical alert, botanical alert, bespoke request, delivery issue, complaint), scores urgency, and auto-escalates medical or botanical safety alerts to a 5-minute SLA before encryption and queueing for a human concierge. A Treasury layer runs automated anomaly detection — ledger balance checks, seal-chain integrity, balance drift, forex exposure — locking withdrawals on any discrepancy. The storefront is an immersive 3D experience (React Three Fiber) with dual checkout: Stripe and native Solana/Ethereum wallets via wagmi and viem. Shipped as a multi-stage Docker image (Node build → nginx serve) built for linux/amd64, matching the AMD judging infrastructure end to end — container platform down to the GPU compute path.

To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits. All hackathon participants will receive on hackathon launch additional compute and API credits for their projects. Ready to build what's next on AMD? Sign up with AMD below to secure your spot.