Nov 2020 – Jul 2023 · 2 yrs 8 mos · Bengaluru, Karnataka, India · Hybrid

• 𝗣𝗵𝗶𝗹𝗶𝗽 𝗠𝗼𝗿𝗿𝗶𝘀 𝗜𝗻𝘁𝗲𝗿𝗻𝗮𝘁𝗶𝗼𝗻𝗮𝗹 – 𝗔𝗜 𝗯𝗮𝘀𝗲𝗱 𝗽𝗿𝗶𝗰𝗶𝗻𝗴 𝘀𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻 𝘁𝗼𝗼𝗹
• Developed a Pricing Simulator to predict SKU sales for PMI’s tobacco portfolio at various levels such as SKU, brand, and manufacturer across key accounts
• Implemented the solution for 15 countries using price elasticities, historical volume trends, and consumer switching maintaining an accuracy of 95%
• Added functionalities like price segment repositioning, SKU delisting/merging, new launch products simulations
• Optimized the backend pipeline to reduce simulation run time by 70% for an enhanced user experience
• 𝗡𝗶𝘃𝗲𝗮 𝗜𝗻𝗱𝗶𝗮: 𝗗𝗲𝗺𝗮𝗻𝗱 𝗦𝗲𝗻𝘀𝗶𝗻𝗴 𝗳𝗼𝗿 𝘀𝗸𝗶𝗻 𝗰𝗮𝗿𝗲 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝘀
• Created a forecast model to predict demand for skin care products across all major cities and channels, including general trade, modern trade, and online sales.
• Collaborated with the team to improve end-to-end project pipelines, including tasks such as data validation, feature engineering, forecasting, post-processing, and model evaluation.
• Utilized a variety of multivariate models, such as XGBoost, SVM, and Random forests, as well as univariate models like ARIMA and exponential moving average.
• Incorporated COVID-19 impact adjustments into the models to ensure accurate predictions.
• Increased the legacy model channel-level accuracy from 85% to 90%.
• 𝗝𝗞 𝗣𝗮𝗽𝗲𝗿 - 𝗣𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝘃𝗲 𝗠𝗼𝗱𝗲𝗹𝗶𝗻𝗴 𝗳𝗼𝗿 𝗣𝗮𝗽𝗲𝗿 𝗠𝗮𝗻𝘂𝗳𝗮𝗰𝘁𝘂𝗿𝗶𝗻𝗴 𝗣𝗿𝗼𝗰𝗲𝘀𝘀 𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻:
• Developed machine learning models for predicting web breaks and detecting anomalies in a paper manufacturing process
• Applied Recursive Feature Elimination with XGBoost, ANN, and Extra Tree Regressor to address imbalanced classification with a 0.4% occurrence rate
• Achieved 85% recall and 30% F1 score for web break prediction

May 2019 – Jul 2020 · 1 yr 2 mos · Bengaluru, Karnataka, India · On-site

• Analyzed patient-level data (prescriptions, diagnoses, procedures, surgeries) for a global healthcare intelligence organization.
• Designed and implemented deductive promotional triggers to target physicians more effectively, based on key diagnosis and treatment events in a patient's journey.
• Analyzed patients' lines of therapy to identify key metrics such as time between treatment lines, physician specialties, and payer metrics.
• Conducted sophisticated persistency and compliance analysis to identify gaps between treatment and patients reinitiating or switching therapy, resulting in improved patient outcomes.

Feb 2018 – May 2019 · 1 yr 3 mos · Bengaluru Area, India · On-site

• 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗲𝗱 𝗠𝗮𝗿𝗸𝗲𝘁 𝗠𝗶𝘅 𝗠𝗼𝗱𝗲𝗹 𝗳𝗼𝗿 𝗛𝗲𝗮𝘃𝘆 𝗠𝗮𝗰𝗵𝗶𝗻𝗲𝗿𝘆 𝗖𝗹𝗶𝗲𝗻𝘁: Developed market mix model using Linear Regression, Feature Selection, and PCA techniques to extract ROI from marketing spending in print and social media channels.
• 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗱 𝗦𝗲𝗴𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗧𝗼𝗼𝗹 𝗳𝗼𝗿 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝗣𝗿𝗼𝗳𝗶𝗹𝗶𝗻𝗴: Created data clustering/segmentation tool on the R-Shiny platform with advanced cluster visualization and customer profiling capabilities, using K-means, Hierarchical Clustering, and DBSCAN techniques.
• 𝗧𝗿𝗮𝗶𝗻𝗲𝗱 𝟯𝟬 𝗕𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗔𝗻𝗮𝗹𝘆𝘀𝘁𝘀 𝗶𝗻 𝗘𝘅𝗰𝗲𝗹: Developed and delivered customized training programs to equip 30 newly recruited business analysts with essential Excel skills and knowledge.

Aug 2017 – Jan 2018 · 5 mos · Bengaluru Area, India · On-site

• Classified the top-performing sales representatives for a cancer drug launched in 20 pilot districts across the United States by providing them visualization capabilities analyzing drug sales and assessing the targeting of health care providers by the sales representatives. This dashboard helped the client increase sales force effectiveness and make better decisions regarding their incentives process for the sales team
• Developed an Excel dashboard for predicting monthly Natural gas usage in a Texas household to calculate the discount on pipeline installation costs for new connections

Jan 2017 – Jun 2017 · 5 mos · Bengaluru Area, India · On-site

• PCIe Verification team:
• 1. Traced the source for register fields having mismatches in values read from the DUT and the mirrored values using Verdi tool
• 2. Executed a shell script to add reset values for register fields having no initial values defined or register fields behaving differently on hard reset and soft reset
• 3. Added register constraints to ensure registers attain only valid values during register randomization using UVM.
• 4. Added register interdependencies in the RAL file to solve register mismatches using UVM

May 2015 – Jul 2015 · 2 mos · Pilani, Rajasthan , India · On-site

• Worked under the mentorship of Dr. Shashikant Sadistap, Scientist 'G', Agri-Electronics Group, Central Electronics Engineering Research Institute, Pilani
• Developed a simple, low-cost, and autonomous Opto-Ultrasonic Sensor System to detect the proportion of water adulteration in milk
• Optical Sensors and Acoustic Sensors were individually tested, modeled, and then developed into autonomous systems using microcontrollers
• Integrated the systems to form an Opto-Ultrasonic Sensor System which displays the percentage of water adulteration in milk samples as well as adulteration by other substances such as urea