• Architected and executed high-performance data pipelines to process massive datasets exceeding 10 billion rows. Leveraged big data technologies Spark and SQL in Databricks to effectively source and compile grocery Ad data for forecasting and analysis. Constructed highly accurate predictive models to predict ad effectiveness for three distinct product categories across two banners. Achieved an outstanding R2 score of 0.85, enabling persuasive storytelling and driving impactful business decisions.

• Developed interactive dashboard that visually showcases diverse customer metrics, like spending patterns across customer types, facilitating in-depth analysis of purchase behaviour and enabling strategic marketing and inventory management. Optimized SQL queries to streamline dashboard data retrieval, reducing the number of queries by over 60%. This optimization led to decreased cloud resource utilization and substantially improved run time.

• Developed a Python and OpenCV-based Machine Learning model for image analysis, enabling automatic table detection and saving. Semi-automated solution drastically reduced manual processing of paper-based bills, saving both time and space by 50%.

• Developed back-end functionalities for the multi-team project ‘Edumonics’, that extracts keyword phrases from text given Django API and implements Natural Language Processing techniques to identify "Subject-Verb-Object (SVO)" combinations.