"MaachaCode team on AMD AI League Hackathon"

Team Idea

We are building an AI-powered retail shelf inspection and compliance platform using AMD-hosted AI inference. The application analyzes shelf photographs to identify products, estimate quantities, and compare observed stock against predefined shelf specifications. It flags potential stockouts, understocking, misplaced products, and uncertain findings. A financial impact module estimates potential margin associated with missing products, while an AI-powered action-note generator provides actionable recommendations for store staff. An interactive Streamlit dashboard brings the entire workflow together. With structured output validation, automated tests, and fallback mechanisms, the system aims to make retail shelf audits faster and more efficient. Since AI-based product counting can be inaccurate due to occlusion and similar packaging, results and financial estimates are clearly marked as unverified and require human validation.

Team Members

Bhavya Aggarwal

bhavyajpg147

Submission

Submission draft in progress