Business Challenge

A leading e-commerce review platform that helps customers make informed purchasing decisions was facing major hurdles as it grew. The platform relied heavily on data from customer reviews, blogs, and forums to recommend the best products. However, with the explosion of data from so many sources, managing and analyzing it became overwhelming. The company was spending too much time and money on manual data review, and its current systems couldn’t keep up with the increasing volume of information. Moreover, they struggled to offer accurate, personalized product recommendations in real time, leaving users without the up-to-date guidance they needed.

Proposed Solution

Recognizing the urgency of these challenges, the platform partnered with Shorthills AI to find a better way to manage their data and improve their product recommendation system. Shorthills AI stepped in with a smart, automated approach. Instead of manually processing all the reviews and feedback from different sources, they set up a system that could automatically collect and sort this information. This data, stored on Amazon’s cloud platform (Amazon S3), was then sent to another advanced system—Azure Databricks—where it was quickly analyzed and organized. Shorthills AI also made sure that the platform’s product recommendations would always stay relevant and accurate by integrating machine learning models that constantly learn and improve. Lastly Shorthills implemented real-time data analysis, allowing the platform to instantly provide product recommendations based on the latest customer feedback.

 Key Outcomes

With Shorthills AI’s solution in place, the platform could now handle over 100 terabytes of data every month, processing what used to take weeks in just a few hours. This massive leap in efficiency not only saved time but also reduced the platform’s operating costs by 30%. More importantly, users noticed a real difference—the accuracy of product recommendations improved by 15%, offering a more personalized shopping experience. Plus, with the ability to handle more traffic without slowing down, the platform was well-prepared for future growth, even during peak shopping periods.

Tech Stack

Azure Databricks, GPT4, Llama, Semantic Search, LangChain