Sellerdashboard.ai

The project commenced in 2018 due to a local retail company’s need for an in-house software solution. The company was previously spending over $5000 monthly on software to manage their Amazon inventory, which was not only costly but also slow and inadequate for their needs. They required a more robust solution tailored specifically to their business requirements, leading to the creation of SellerDashboard.ai.

Initially, SellerDashboard.ai was merely a research tool designed to analyze a list of products and determine which would sell best on Amazon. This simple concept presented several challenges. First, I had to navigate the MWS API to fetch product information, including sales rank. Then, I developed an algorithm to estimate monthly sales based on the sales rank. Another hurdle was adapting this functionality to work within the constraints of the MWS API, which limits frequent usage.

As the project expanded, it evolved into a comprehensive warehouse management suite. The main challenge here involved refining the pick and pack process and managing inventory flow with utmost precision.

Another complex aspect of the project was repricing. On Amazon, competitive pricing is crucial for winning the buy box. However, simply lowering prices isn’t always effective. There’s a delicate balance between price and fulfillment method. For FBA sellers, it’s possible to maintain a higher price point—about 0.1 or 0.2 above the lowest offer—while still achieving a high buy box win rate. I developed a strategy that integrates this pricing model into the Laravel framework to adjust prices on Amazon via API. With the aid of RabbitMQ, which facilitates multiple executions during pricing events, the code I wrote responds instantly to pricing signals, resulting in a 2% improvement in margins and maintaining an 80% buy box win rate.

Finally, SellerDashboard.ai now features a customized FBA workflow that reduces preparation time by 20%. This efficiency was achieved by closely analyzing and understanding how the warehouse team operates. By applying AI & ML, I learned which box sizes are best suited for specific products and quantities, enabling optimal box selection and quantity counting. This knowledge significantly speeds up the packaging process for the team.