Feb 2021 – Mar 2023 · 2 yrs 1 mo · Gurugram, Haryana, India · On-site

• At Zomato-Blinkit, I've revolutionized retail efficiency and customer experience through advanced data science initiatives:
• Optimized ML-Driven Demand Forecasting:
• Engineered a forecasting system across 304 stores, managing daily replenishment of 3.3M items.
• Achieved $9.5M annual savings, boosting product availability to 95-98%.
• Utilized Temporal Fusion Transformers (pytorch Forcasting) and Facebook-Prophet, enhancing forecast accuracy by up to 25%.
• Implemented LGBM and Catboost models, refining forecast precision for 3.2M low-selling items.
• Innovative Product Recommendation & Market Analysis:
• Developed a Hybrid Collaborative Filtering System, achieving an AUROC of 87%.
• Identified 1000 high-confidence association rules through Market Basket Analysis.
• Integrated Large Language Models, optimizing association rule generation by 13%.
• Crafted a custom NLP dictionary for product name correction, attaining 93% accuracy.
• My role at Zomato-Blinkit has been pivotal in leveraging data science to drive significant business improvements and enhance customer engagement.

May 2020 – Jul 2020 · 2 mos · Worli, Maharashtra, India

• Created propensity model via supervised learning techniques, used to create target groups for cross-selling marketing campaign, churn prediction models, Recommendation Engine

Apr 2020 – Jul 2020 · 3 mos · Mumbai, Maharashtra, India

• Project 1: Product recommendation system and market basket analysis
• Project 2: Binary Classification Model For Uninstallation Of The Application (Churn prediction)
• Project 3: Fraud detection at store level

Jan 2020 – Feb 2021 · 1 yr 1 mo · Bengaluru, Karnataka, India · On-site

• Time-series Forecasting and Anomaly Detection
• During my time at CRED, I was instrumental in developing 'Eagle', an advanced fintech anomaly detection and resolution system.
• Key Contributions:
• Robust Anomaly Detection Pipeline: Engineered a pipeline monitoring over 1000 fintech metrics, especially during India's premier cricket event. Successfully identified critical anomalies in real-time.
• Significant Financial Impact: Uncovered and resolved a credit card referral abuse trend and installation to signup failures, safeguarding the organization from a potential loss of $500K in just one week.
• Enhanced Anomaly Resolution: Formulated custom statistical algorithms to pinpoint anomaly causes, achieving a staggering 1000% improvement in resolution efficiency.
• Forecasting with Facebook's Prophet: Adapted Facebook's Prophet for optimal forecasting, enhancing average metric accuracy by 20% through tailored uncertainty intervals.
• Innovative Dashboard Creation: Developed a comprehensive dashboard to display anomalies with weighted importance, offering insights into potential causes and resolutions at both metric and feature levels.
• My tenure at CRED was marked by significant contributions to enhancing financial security and operational efficiency, underscoring my capabilities in data science and anomaly detection.

Jan 2020 – May 2020 · 4 mos · Gurugram, Haryana, India

• Worked on Email bot, Big data, dynamic discounting, Elastic search, Django, Flask, AWS Sagemaker, S3, RESTApi's, Docker, Hadoop

Dec 2019 – Jan 2020 · 1 mo · Gurugram, Haryana, India

• Worked on building customer retention models and Dynamic Discounting models

May 2019 – Jul 2019 · 2 mos · Kharagpur Area, India

• Developed an algorithm(MORE Algo) to mine optimum association rules to detect the values BP/BMI/CH for the patient's, without actually performing actual tests :)

Jul 2018 – Apr 2019 · 9 mos · Kharagpur Area, India

• Led a team of 32 associate members to ensure the effective and smooth functioning of the organization and its work.
• Involved in on-campus publicity and event management of two-day events like Paintball, Zip-line, Rappelling, zorbing, etc.
• Part of a video editing team and prepared several videos for the treks, events and introductory seminar attended by 8000+ students.