The client, a leading Fintech company, provides Open-To-Buy (OTB) credit services for Lease-To-Own (LTO) Software-as-a-Service (SaaS) offerings. They approached codvo.ai with a business challenge of developing an NLP model that accurately classifies retail products for Lease-To-Own services during checkout on Amazon.
The client's main business challenge was the inability to accurately identify leasable products during checkout on Amazon, which resulted in delays in decision-making and increased the risk of offering credit on non-leasable products. This impacted the client's ability to provide an optimal customer experience and affected their bottom line.
Additionally, the absence of a mechanism for product classification led to sub-optimal utilization of their Open-To-Buy credit services. The client needed a solution that would enable them to accurately identify leasable products and provide a seamless shopping experience to their customers.
Our approach to the client's challenge involved developing an NLP model based on the LSTM algorithm to accurately classify retail products as leasable or non-leasable during checkout on Amazon. We utilized Python Scikit Learn and Jupyter Notebook for model development and Docker for containerization, ensuring that the model was portable and could be easily deployed.
We used AWS services such as S3 for data storage, Lambda for serverless computing, and SageMaker for model training and deployment. We also created a scalable API using AWS API Gateway to enable seamless integration with the client's LTO partner platforms. The API enabled easy management and deployment of the NLP model, ensuring that the solution was reliable and scalable.
Our approach focused on building a robust and accurate NLP model that could reliably identify leasable products during checkout on Amazon. We integrated the model with the client's LTO partner platforms, enabling end-users to accurately identify leasable products during the checkout process. This integration ensured a seamless and hassle-free shopping experience for customers and enabled the client to provide better OTB credit services to their customers.
.png)
The tech stack for this project included Python ScikitLearn, Jupyter Notebook, Docker, GitHub, AWS S3, AWS Lambda, and AWS SageMaker.
-Achieved 85% classification accuracy for identifying leasable products, significantly reducing errors during checkout
-Improved utilization of Open-To-Buy credit services, leading to better credit decisioning and increased conversion rates
-Enabled faster and more accurate checkout experiences, enhancing overall customer satisfaction
-Minimized credit risk by preventing financing of non-leasable items, protecting revenue and compliance
-Delivered a scalable and easily deployable NLP solution integrated with LTO partner platforms via AWS API Gateway
-Reduced operational overhead by automating product classification, eliminating manual reviews and delays
