by Komal Vachhani
See all projects here!
https://www.youtube.com/watch?v=TQyT38PBkbE
ProductPal Proposal Slide Deck
ProductPal Proposal Slide Deck
💡 Idea
The idea behind ProductPal was to create a shopping companion that would assist users in finding relevant items based on their needs. The goal was to develop a tool that would analyze website content, provide personalized suggestions, and offer a simple and user-friendly interface for seamless navigation using GPT.
🔨 Tools
- Dynamic Javascript Scripts: Dynamic Javascript scripts were employed in the backend to continuously monitor and relay the most up-to-date window data. This allowed for real-time analysis of website content and ensured that the suggestions provided were relevant and accurate.
- GPT-3 (Davinci Text Model): The project leveraged OpenAI's GPT-3 Davinci Text model, a powerful language model, for natural language processing (NLP) tasks. GPT-3 was used to process and interpret the website content, enabling the system to generate the best suggestions for users based on their needs.
- Chrome extension: The frontend interface of ProductPal featured a simple popup screen that provided users with all the relevant suggestions for similar products. This intuitive and unobtrusive design allowed for a seamless user experience.
- Website Navigation: Users were able to easily navigate to suggested websites directly from within the popup screen. This feature provided convenience and efficiency, enabling users to explore and compare recommended products effortlessly.
🔑 Key Learnings/Takeaways
- Real-time Data Analysis
- Natural Language Processing
- User-Centric Design
- Personalized Recommendations
- Seamless Integration