Sep 2022 – Aug 2025 · 2 yrs 11 mos

Oct 2020 – Sep 2022 · 1 yr 11 mos · Bengaluru, Karnataka, India

• ◦ Developed a predictive model using Random Forest to predict if the invoices will be paid early, on time or delayed, with an Accuracy of 94%, integrated results in Qlikview
• Tools/Algorithms – Python, PySpark, Random Forest Classifier
• ◦ Reliable algorithm for invoices that will be posted and cleared in the upcoming month along with forecasting how much overdue invoices will be cleared in the next month and helped Qlik developer in integrating the results in QlikView Dashboard. Improved Accuracy to 97%
• Tools/Algorithms – Python, Databricks, Time Series Forecasting, ARIMA
• ◦ Forecasting based on statistical models and handshake with the business on accounts payables forecast. Used open invoices and clearings data worth of $12B (yearly) on the world level to forecast CashFlow for the next 6 months along with 19 different markets. Run 30 iterations of different models before concluding the forecasting results, achieved an Accuracy of 95% with less than $50M deviation in CashFlow quarterly Tools/Algorithms – Python, Databricks, Time Series Forecasting, Auto ARIMA, ARIMA, SARIMA
• ◦ Built a model to forecast the net CashFlow of seven different payment components, reduced MAPE to 5% Tools/Algorithms – Python, Databricks, ARIMA
• ◦ Implemented a Fuzzy Matching Tool to reliably link agencies and employers and provided a confidence score on every match attempt. Correctly mapped 550k employers 95% of total employers with respective agencies
• Tools/Algorithms – Python, PySpark, Databricks, Fuzzy Matching, Levenshtein Distance

Apr 2019 – Oct 2020 · 1 yr 6 mos · Bengaluru Area, India

• ◦ Used Genetic Algorithm-optimized feature list to predict adherence. Analyze patient cohort and identify the drivers and barriers of medication adherence, augmented Recall of 75% and Up to 85% prediction accuracy for smaller patient cohorts. Presented at PAN India conference
• Tools/Algorithms – Python, PySpark, MySQL, AWS, EC2 Server, Random Forest, XGBoost
• ◦ Designed a predictive model to identify future potential writers and drops to optimized touch-point interactions with RMSE of 4.1 and suggested key-drivers for sales, integrated findings in R Shiny app
• Tools/Algorithms – R, R Shiny , Auto-Encoder, XGBoost Classifier
• ◦ Established a pipeline to identify and characterize key drivers that drive patient adoption of 2 different products and ECP behaviour as well to inform targeting and sales strategy using PU Learning
• Tools/Algorithms – Python, Impala, HUE, MySQL, AWS, EC2 Server, Shapely, PU Learning
• ◦ Developed an Ensemble Tree based Boosting model to identify factors that are associated with patients starting / switching to an IV dosage
• Tools/Algorithms – Python, MySQL, Databricks, Ensemble, Random Forest, XGBoost

Sep 2018 – Nov 2018 · 2 mos

• Developed a Job Description parsing tool that can map the requirement to the key terms in the database and provide insights on patterns of the job requirement.

Jun 2018 – Mar 2019 · 9 mos

• Built a hybrid recommender system (Collaborative Filtering + Content-based techniques) that give a list of recommended jobs to the user in the Job Portal.
• Developed an ensemble of SVM, Logistic Regression and XGBoost for dataset provided to determine who gets shortlisted and approved by employers. Achieved an Accuracy of ~68% Recall of 71%.

Nov 2017 – Mar 2021 · 3 yrs 4 mos

• I founded the Data Science Club, which fostered interdisciplinary research on applications of Data Science across departments. I, along with the core members, laid down a diligently worked out road map for the organization.

Oct 2017 – Nov 2017 · 1 mo · Remote

• Worked on Predictive model using LGBM to predict the probability of successful calls initiated by the MS which helps in evaluating the network performance.

Dec 2015 – Jan 2016 · 1 mo · Jaipur Area, India

• In the course of this internship, I had made some changes in the Book Returning Rules using PHP.

May 2015 – Jun 2015 · 1 mo · Jaipur Area, India

• The internship involved working on the Management of Customer reviews.