ShelfPulse AI

Overview:

ShelfPulse AI is a real-time retail intelligence system that detects empty shelves, tracks inventory, and predicts restocking needs — helping stores eliminate revenue loss due to stockouts.

Turning shelves into data-driven decision systems

---

Key Features:

Real-Time Detection – Live camera-based shelf monitoring

SKU Recognition – Identify products at item level

Empty Slot Alerts – Detect out-of-stock instantly

Smart Dashboard – Visual stock and performance insights

Predictive Replenishment – Forecast demand and restocking

---

How It Works:

Camera Feed → YOLOv8 Detection → SKU & Empty Slot Analysis → Dashboard → Forecasting → Restock Recommendations

---

Tech Stack:

Frontend: React, Tailwind CSS

Backend: FastAPI / Node.js

AI/ML: YOLOv8, ONNX Runtime

Analytics: Prophet Forecasting

Database: PostgreSQL / Neon

---

Project Structure:

ShelfPulse-AI
 ┣ models        → Trained and exported models  
 ┣ frontend      → React dashboard  
 ┣ backend       → APIs and forecasting  
 ┣ data          → Dataset and annotations  
 ┗ README.md  

---

Impact:

Reduce out-of-stock losses

Improve operational efficiency

Increase retail revenue

Enable smarter decision-making

---

Mindmaps:

Tech Stack

---

Why It Stands Out:

Combines Computer Vision and Forecasting

Works in real-time with low latency

Scalable for multi-store retail chains

Targets a large global retail problem

---

Future Scope:

Mobile app for store managers

Multi-store analytics dashboard

Advanced AI demand prediction

IoT and smart shelf integration

---

Team:

• Priyanshi Shah (Leader)
• Shlok Shah
• Kavya Shree
• Bhagya Shah
• Sarvesh Chhatani

---