• Enhanced responsiveness by 10%, quantifying complexity of cases using Linear Regression; Built and fine-tuned OpenAI model, determining and extracting uncorrelated variables from case summaries.• Aided a Retired Player organization in accessing the quality of life of former players through comprehensive survey to gauge NPS, Satisfaction and Wellness Score; Additionally, implementation of statistical methods to quantitatively measure health and financial well-being, providing actionable… Show more • Enhanced responsiveness by 10%, quantifying complexity of cases using Linear Regression; Built and fine-tuned OpenAI model, determining and extracting uncorrelated variables from case summaries.• Aided a Retired Player organization in accessing the quality of life of former players through comprehensive survey to gauge NPS, Satisfaction and Wellness Score; Additionally, implementation of statistical methods to quantitatively measure health and financial well-being, providing actionable insights for program enhancements.• Delivered and presented 15+ Tableau dashboards within 2-month project timeline for Marketing and Ticketing Revenue analysis; Designed wireframes in Figma, cleaned and aggregated over 30+ data sources using Snowflake for dashboard development; Led a team of 2 developers, understanding business requirements and effectively showcasing results to the executives.• Optimized resource allocation across 64+ attributes affecting fan’s gameday experience using Survey Analysis Framework; Developed Impact Analysis method to measure attribute impact on overall experience, employing Gradient Boosting Regressor to quantify and debias attribute ratings, enhancing accuracy of resource allocations decisions; Constructed Power BI dashboard, delivering results to 32 clubs for performance evaluation and comparison with benchmark.• Applied topic modelling and sentiment analysis on fan’s feedback, extracted hidden themes and topics for informed business decisions.• Managed a team of 3 to deliver data-driven insights to stakeholders across International TV, Network and Social Media Platforms; Interpreted business requirements, extracted data from 20+ SQL tables, and produced analytical reports highlighting key insights; Assessed and compared TV deals valuation of competitors, conducted in-depth viewership trend analysis, and utilized Machine Learning model to reduce competitive data procurement costs by $15K. Show less

• Extracted, collected, and filtered geo-tagged images of giraffe using Platform API (Twitter, Flickr, Instagram, etc.)• Built a Regression model to observe the migration pattern of giraffe using geo-spatial visualization

Courses: Introduction to Computers Programming and Data Science on Ramp (mini courses include Data Visualization using Tableau, Machine Learning, Feature Selection, Web Scraping and NLP in Python)

• Applied advance transformation and data modelling techniques using OpenCV to analyze tissue image data• Experimented, designed, and fine-tuned CNN object detection architectures like Faster RCNN, Mask RCNN, and AlexNet to accurately detect functional tissue unit on different stained tissue blocks• Optimized and evaluated the performance on accuracy metrics like F-score, Precision, Recall and IoU, attained 8% higher score than existing models

• Increased 40% efficiency of weekly advertisement financial reporting, engineered ETL pipeline using Google CloudPlatform.• Implemented automated system integrating Python with Google Sheets to deliver weekly financial reports to stakeholders.

• Devised a probabilistic model using Gradient Boosting Classifier to predict propensity of customers to buy personal loans which optimized marketing cost by 30% with a coverage of 64% of total customer base • Collaborated with a leading FMCG client for voice-based ordering system in the sales rep application; Extracted keywords like product name, quantity and measurement from Google speech API’s output using nltk

• Created time series model using different models like Prophet, ARIMA and STL modelling to forecast different network’s parameters and compare actuals vs forecast• Segmented source IP addresses on network’s various parameters like bytes, packets, session time, denied flows, etc. using K-Means Clustering to flag anomalous traffic• Developed a multi-class classification model using Random Forest to predict potential cross-sell customers which achieved 74% of target cross-sell customers… Show more • Created time series model using different models like Prophet, ARIMA and STL modelling to forecast different network’s parameters and compare actuals vs forecast• Segmented source IP addresses on network’s various parameters like bytes, packets, session time, denied flows, etc. using K-Means Clustering to flag anomalous traffic• Developed a multi-class classification model using Random Forest to predict potential cross-sell customers which achieved 74% of target cross-sell customers in 34% of the total customer base Show less

Designed mechanism for probing micron movement level resolution for Electrolithography Machine | Devised new ideas for making Smart Urinals which can be installed in public toilets

In the Inbound team, pitched multiple deals to the service providers, converting 120+ service providers| Received Letter of Recommendation from Abhiraj Bahl (Founder of UC) for acquiring 200% of the targets

Responsible for examining the Quality and analysing MTM-UAS Time study of various delibra in Body Shop,Car Division thereby increasing efficiency of workers by 25%