Created for the Google Cloud Vertex AI hackathon.
Introducing Summazon: Revolutionizing Amazon Review Summaries.
In a vast sea of products, navigating through a canopy of Amazon reviews can be overwhelming. That's where Summazon comes in. Our cutting-edge AI-powered solution revolutionizes the way Amazon reviews are summarized, offering a comprehensive and insightful analysis of product feedback.
Summazon first splits reviews into distinct categories, such as pros and cons, and uses those reviews to produce assessments on four key metrics: price, performance, reliability, and quality. It then takes this categorized review data and generates concise and informative summaries. Ultimately, allowing users to save lots of valuable time and effort, providing them with the essential insights they need to make well-informed purchasing decisions.
Whether you're a tech enthusiast searching for the best-performing gadgets, a budget-conscious buyer comparing prices, or a company seeking product feedback analysis, Summazon is your go-to tool for efficient and insightful review assessments.
Now it’s time to say goodbye to information overload and let Summazon help you rise above the canopy of Amazon reviews.
Frontend: The front end was constructed using Next.js and Tailwind CSS, with hosting on Vercel. Every time a search is performed on the page, regex is used to extract the product ID from the Amazon Link, and an API request is made to our backend.
Backend: The back end was constructed using Flask. When a request is made, Amazon review data is scraped using Playwright. This scraped review data is sent along with our prompts to the Google PaLM-2 Chat-Bison 001 AI (a Large Language Model similar to ChatGPT), for summaries.
- Connecting the front-end and back-end posed significant challenges, with numerous networking and CORS errors. Eventually, we resorted to using a CORS proxy along with ngrok to enable front-end access to the back-end. We recognized the need to gain more experience with server-side requests.
- Due to the nature of web scraping and AI processing, search requests often experienced significant delays. To mitigate this, we implemented request caching, although further improvement in faster web scraping techniques will be necessary.
This project introduced us to several technologies we hadn't used before, such as Next.js and the Google Cloud API. We gained valuable experience in incorporating these technologies into our project.
Our website successfully leverages Google Cloud Vertex AI to provide users with easy access to organized Amazon review summaries. We take pride in offering a clutter-free and user-friendly format, enabling users to quickly find the reviews they need.
At Summazon, we aim to continually enhance our service, always looking for avenues for our AI-powered tool to provide more value to our users. One of the significant benefits we aim to include in our feature pipeline is the ability to provide product comparisons.
Product comparison is an essential aspect of shopping, whether online or offline. Often, consumers are torn between two or three similar products, and making the right choice becomes a pain point. With the planned product comparison feature, Summazon aims to not just summarize reviews but to make these summaries comparative. This will provide consumers with a platform where they can easily compare similar products based on the core metrics like price, performance, reliability, and quality.
Furthermore, we aim to extend our service to incorporate more than just Amazon products. Other platforms, such as eBay, Best Buy, Walmart, and more, also host a plethora of product reviews that consumers need help navigating.
As part of this expansion, we envisage launching browser extensions for closer integration with these various shopping platforms. This implementation will provide a more seamless transition from browsing product reviews to making that final purchase, enhancing user experiences and making online shopping even more convenient.
Finally, we have the ambition of broadening our horizons beyond mere product review summaries. We hope that our algorithm, in time, can be tailored to accurately detect market trends based on consumer reviews. By analyzing the tone and content of reviews, we believe we can provide insight into what products are trending, what features consumers are prioritizing, and even the potential emergence of new market segments.
This is a Next.js project bootstrapped with create-next-app.
First, run the development server:
npm run dev
# or
yarn dev
# or
pnpm devOpen http://localhost:3000 with your browser to see the result.
You can start editing the page by modifying pages/index.tsx. The page auto-updates as you edit the file.
API routes can be accessed on http://localhost:3000/api/hello. This endpoint can be edited in pages/api/hello.ts.
The pages/api directory is mapped to /api/*. Files in this directory are treated as API routes instead of React pages.
This project uses next/font to automatically optimize and load Inter, a custom Google Font.
To learn more about Next.js, take a look at the following resources:
- Next.js Documentation - learn about Next.js features and API.
- Learn Next.js - an interactive Next.js tutorial.
You can check out the Next.js GitHub repository - your feedback and contributions are welcome!
The easiest way to deploy your Next.js app is to use the Vercel Platform from the creators of Next.js.
Check out our Next.js deployment documentation for more details.