We wanted to empower shop owners with deeper insights into their customers’ behavior, going beyond traditional metrics like inventory and sales. By analyzing in-store activity, our goal is to provide actionable data that helps optimize store layouts, improve customer experiences, and ultimately boost sales.
Shoplytics leverages live video feeds (e.g., store CCTV cameras) to provide real-time analytics for shop owners. The app detects the number of customers in the store and pinpoints their locations. It generates valuable insights, such as:
- The most popular shopping times.
- Heatmaps showcasing the most visited areas in the store.
- Individual customer paths throughout the store
- Popular customer visiting times
Customer Detection: Utilized Ultranalytics to accurately detect and track customers. Backend: Powered by Flask, ensuring seamless data processing and API integration. Frontend: Built with React, delivering an intuitive and interactive user interface Challenges we ran into
- Analyzing the data to convert into a heatmap
- Accomplishments that we're proud of
- Getting the live video feed and customer detection working
We plan to take Shoplytics to the next level by integrating a Large Language Model (LLM) to:
- Analyze the collected data.
- Provide actionable recommendations to enhance the shopping experience.
- Suggest ways to optimize store layouts and drive sales growth.
In the backend directory:
pip install requirements.txt
Run the flask server:
py server.py
Run the web app:
npm start