🔍 Overview
We've developed a web application using Streamlit, a Python-based frontend UI framework, incorporating advanced AI technologies like CrewAI for multi-agent systems and Langchain to leverage the capabilities of Gemini-1.5-pro and LLaMA-70B.
🤖 AI Integration
The system integrates Large Language Models, LLaMA-70B and Gemini-1.5-pro, functioning as multi-agent AI systems that autonomously retrieve and analyze information from the internet to provide optimal product recommendations. These models do not require specific dataset training, as they fetch necessary information dynamically using various tools.
⚡ Performance
The application uses Groq API for LPU inference, significantly reducing model response time from minutes to milliseconds.
👥 Agents
It features three simultaneously operating agents:
- Product Recommendation Agent: Suggests the best brands and varieties.
- Image Question-Answering Model: Analyzes images and answers queries related to objects within them.
- Web Search Agent: Scours the web for products, offering relevant links and approximate prices tailored to the user's country and currency.
- Streamlit 🖥️
- Langchain 🔗
- CrewAI 🤝
- Groq API 🚀
🔧 User Profiles and Preferences
Incorporate user profiles that capture individual preferences, purchase history, and browsing behavior. This could refine recommendations and responses, making them more relevant to each user.
🔄 Adaptive Learning
Implement mechanisms where the system learns from user interactions to improve recommendations over time. This could involve reinforcement learning or other adaptive techniques.
- Product Recommendations: Offer personalized product recommendations based on preferences, browsing history, and current trends.
- Price Comparison: Provide comparative pricing and product availability across different e-commerce platforms.
- Visual Search: Allow users to upload images to find similar products or detailed information about items.
- Virtual Shopping Assistant: Act as a virtual assistant that answers questions, suggests products, and helps users navigate online stores.
- Automated Query Handling: Manage customer inquiries and support requests by analyzing and responding to questions in real-time.
- Country and Currency Adaptation: Offer product recommendations and pricing tailored to the user’s country and currency.
- Regional Trends: Incorporate regional trends and preferences to tailor recommendations to local markets.
# 1. Create the following api keys and include them in .env file
GEMINI_API_KEY='YOUR API KEY HERE'
GROQ_API_KEY='YOUR API KEY HERE'# 2. To install the required dependencies
pip install -r requirements.txt# 3. To run the python file using streamlit
streamlit run main.py
