- Preparing training data for the Machine Learning model by annotating the essential attributes in the 250+ documents.- Utilized the Langchain framework to train a machine-learning model on HSN documents. This implementation significantly enhanced the accuracy of predicting HSN codes based on product descriptions. As a result, it streamlined the manual process of code identification and optimized efficiency in code assignment.- I engaged in fine-tuning of OpenAI model (GPT-3.5) utilizing the HSN dataset.- Created an EwayBill module using HTML, CSS and angular to streamline and automate filling out the E-waybills(bill required for transporting shipments) form, reducing the manual workload.

Analyzed customer data and identified clusters of locations that could be efficiently served by a single delivery route from the cluster's centroid, resulting in an average reduction in distance travelled per route.Implemented a novel clustering approach using the Travelling Salesman Problem to optimize delivery routes within defined thresholds which can result in the reduction of delivery costs.