Ensemble Model for Predictive Maintenance- Conducted a deep-dive analysis of various ML models which were created using Vibrations and Acoustic Data that predicts fan failure of an Air Handling Unit. Combined the individual models to form an Ensemble Prediction model, which is scalable and provides an increased accuracy than individual models.Optimization of Supplier Risk Metrics- Developed a feature selection framework using PCA to identify important features out of over 30 features contributing to supplier risk, leading to an increase in the accuracy by over 10%.KPI Extraction using NLP- Worked on extracting KPIs such as Stockout Date and Operations Risk from manually entered comments in the system through NLP techniques. This helped the stakeholders by reducing the manual efforts required to update the KPIs.

Process Automations- Formulated process flows for automation of multiple Supply Chain reports. This heavily reduced the daily manual efforts involved in refresh process.

Predictive Maintenance for Aircrafts- Led a team of four to interpret log pages generated during aircraft maintenance using N-grams in R and translated findings into a visual tool which was leveraged by maintenance engineers to classify open log pages as high/low riskAircraft Wi-Fi Connectivity Risk Analysis- Conceptualized and built Wi-Fi connectivity metric at individual flight level using interpolated data and visualized results into a Spotfire-based global dashboard- This was used by stakeholders to understand risks at granular level which helped them in business conversations with Wi-Fi vendorsGlobal Solution for KPI Visualization- Developed backend data structure using Python for a tool aiding end to end KPI visibility of the supply chain for a Fortune-100 CPG companyTrade Promotion Optimization Framework- Computed impact of various factors on sales volume of a Fortune-100 CPG company using R and provided optimized calendar which helped business to strategically plan upcoming promotions

Driver KPI Metrics Identification- Applied various regression models in R to understand KPIs which drives B2B sales of a technology giant- These results enabled business stakeholders to estimate changes required in key drivers to boost salesDemand Forecasting- Formulated and built time-series forecasting framework that provides week-level demand forecasts for a prominent mobile phone assembler at multiple markets