In the summer of 2024, I had the privilege of interning as a Data Science Intern at Kenvue, Bengaluru. This experience allowed me to dive deep into the intersection of data science and supply chain management, where small improvements in forecasting can lead to massive real-world impact.
The Challenge: Accurate Demand Forecasting
Kenvue, as a global consumer health leader, deals with thousands of SKUs across markets like the US and Canada. Accurate demand forecasting is critical for ensuring product availability, optimizing inventory, and reducing costs. My task was to enhance forecasting accuracy while also making insights more accessible to cross-functional teams.
Optimizing Models for Real-World Impact
I started with existing demand forecasting models and worked on improving them through hyperparameter tuning and feature optimization. These efforts resulted in measurable improvements:\ - 2--3% accuracy gain in Canada\ - 1--2% accuracy gain in the US
While these numbers may look modest, in the context of large-scale supply chain operations, they can translate into millions in savings and better customer satisfaction.
Classifying Demand for Smarter Planning
Beyond forecasting, I also worked on demand classification. By applying ABC analysis and segmenting over 1,000+ SKUs into demand patterns like Smooth, Intermittent, Erratic, and Lumpy, I helped create a framework for smarter planning strategies. This allowed supply chain teams to prioritize resources and respond better to market variability.
Making Data Accessible with Dashboards
To ensure that insights didn't stay locked in models, I built interactive Excel dashboards with pivot tables and charts. These dashboards increased clarity and usability for cross-functional teams by upto 50%, helping business users engage directly with data-driven insights.
Lessons Beyond the Numbers
This internship taught me more than just data science techniques. Some of the key learnings were:\ - Business context matters: Even the most accurate model is useless unless it addresses real business challenges.\ - Communication is key: Visualizations and dashboards are as important as the models themselves.\ - Supply chain is dynamic: Forecasting isn't about perfection---it's about building systems that adapt to uncertainty.
Looking Ahead
My internship at Kenvue gave me valuable exposure to the practical applications of data science in supply chain management. It showed me how technical skills, when combined with domain knowledge, can deliver outcomes that truly matter.
I am grateful to my mentors and teammates at Kenvue for their guidance and collaboration. This experience strengthened my passion for applying data science to solve impactful, real-world problems.