Arun prasadh Senthil

Master's

Data Science

Accenture Solutions

Data Engineer/ML Engineer

- Nike MANA - Created Power BI dashboards to track inventory and forecast sales, and another for understanding geographic sales analysis. These helped in making better informed decisions to manage stock and target markets effectively - Implemented a data pipeline on AWS Glue using PySpark to enhance supply chain efficiency by analyzing and optimizing inventory levels, order fulfillment, and delivery timelines - Employed regression analysis and machine learning algorithms in Python to forecast demand patterns for a retail company. Utilized Google Cloud's AI Platform for model training and deployment. Grouped customers based on their historical purchasing behavior and preferences. Implemented dynamic pricing strategies to optimize revenue and inventory management, leading to increased sales and improved profit margins - Developed internal automation workflows and a user-friendly website using Power Apps to streamline and enhance team engagement activities - Provided mentorship for associates and guided the team during development and deployment of pipelines - Participated in the Solarathon Hackathon with three international students, where we collectively built an FAQ generator using Langchain, OpenAI functions and deployed it in the Ploomber environment with dicord server serving as the data source - Nike SEC - Implemented a suite of chatbots at Nike, leveraging NLTK and Spacy libraries, to assist users with a range of services, from providing real-time updates on order status and offering fitness guidance - Built dashboards using PowerBI to monitor Stories and task for agile development which provided detailed insights to allocate resources in future cycles - Engineered a streamlined data pipeline using PySpark and Pandas that collects and consolidates customer data, covering demographics, usage history, interactions, and other pertinent attributes. The data undergoes meticulous preprocessing in PostgreSQL, where missing values are addressed, irrelevant features are pruned, and the entire dataset is transformed into a structured format, ready for comprehensive analysis - Developed testing framework to validate migrated data from Hadoop to Databricks - Built pipeline using Google cloud composer and Apache beam to identify product performance and revenue trends and scheduled the flow to generate timely reports to stakeholders