Case Study: Amazon Product Rate Prediction

OVERVIEW

To build an AI engine that would automate the purchases at US purchase and predict competitor rates at India for the list of products to be sold by the merchant.

Challenges

• Scrape multiple market places in US including competitors data to gain information on the products’ selling price.
• Scrape multiple discounts and coupons websites.
• Scrape multiple products with competitors, product history, discounts, coupons and match inventory to predict the selling price.

Solution

• Very high complex rules based engine with multiple market places data compared and applying discounts based timeline.
• Compare inventories in hand to order the product from the respective market places when inventories are low.
• Predict the competitors price and notify the admin on the competitive price to be sold.
• Multiple scheduling jobs to continuously gather the data from various market places to keep the engine updated.

Technology Used

• Server – Python, .NET