Teaching

**Projects And Achievements** Throughout my career, I have successfully contributed to numerous projects deliveringinnovative solutions that have had a significant impact on business outcomes. My commitment to excellence and my ability to adapt to emerging technologies make me avaluable asset in any data-driven project. **Advanced Pipeline Leak Detection System** Associated with LTI - Larsen & Toubro InfotechDeveloped a powerful pipeline leak detection system with a multi-sensor datafusion approach. The system uses a complicated logic flow algorithm to analyze readings from numerous embedded sensors and detect anomalies that indicateleaks. The next phase will involve installing predictive machine learning models on Azure cloud, data warehousing in Snowflake, improving early leak detectioncapabilities, and overall pipeline safety. ❖ **ESG Manufacturing - Performance Analysis and Improvement Tool** Associated with Soothsayer Analytics The ESG Performance Analysis and Improvement Tool is a sophisticated AI-powered solution for complete environmental, social, and governance (ESG) data analysis. It collects organizational data from a variety of sources, including Marketpsych, Refinitiv, Bloomberg, and corporate websites, and uses powerful machine learning, generative AI, and statistical approaches on AWS. The programme examines ESG measures, compares firm performance to the ESG league, and generates thorough reports that emphasize critical facts, performance gaps, and ideas for improvement. ❖ **FoxDox for Foxen** Associated with Soothsayer Analytics Foxen is having difficulty obtaining essential data from various insurance documents such as COIs. To address this, we created an intelligent AI-driven document processing system. This system uses a Document Type Identifier to automatically classify documents, extracts key information from various COI documents using a refined LLM-based data extractor, validates the extracted data against a database using a Data Validator, and finally accepts or rejects the documents before storing the results in the database. ❖ **Opportunities Forecast for Belrono** Associated with Soothsayer Analytics Belron, a global leader in car glass repair, replacement, and recalibration, is facing operational issues due to unexpected demand and changing business cycles. To solve these difficulties, we created a machine learning (ML) application. Powered by a powerful data engineering pipeline, the application uses past sales, weather, mobility, and advertising data to provide granular demand estimates for the next three months at both the national and district levels. Time Series Clustering algorithms are used to group districts that share similar historical patterns and variables. ❖ **Plant Residue Prediction for Efficient Pre-Sales Operations** Associated with LTI - Larsen & Toubro Infotech Created a sophisticated machine learning (ML) system that forecasts plant residue by combining IoT data from OSIPI meters and SCADA with a predictive analytics framework. used machine learning techniques and sophisticated statistical models to achieve an error rate of 6%, which was lower than the standard of 9% for optimizers. Future improvements could include contextual data integration to better forecast accuracy and NLP-driven analysis of team conversations to reveal operational insights that are concealed. ❖ **Pump Failure Detection and Proactive Diagnosis System with** Cloud-Based Analytics Associated with LTI - Larsen & Toubro Infotech Using Azure services, we implemented a cloud-based predictive maintenance system for centrifugal pumps. The system uses machine learning to detect anomalies and classify statuses while processing data from various pump stations. This strategy greatly decreases production downtime and increases maintenance intervals. The model's scalability and versatility are ensured by its modular architecture, which allows for easy expansion to more pump stations. ❖ **Strategic Backorder Prediction Model** Developed a thorough backorder prediction model combining real-time analytics with historical data. Based on different operational factors, the model accurately anticipates backorders by using SMOTE to balance the extremely skewed dataset. Supported by a strong data processing pipeline, this strategic tool helps with customer happiness and inventory management. ❖ **Tailored Customer Engagement with Advanced CRM Analytics** We have developed a cutting-edge platform that combines SAP CRM data with Big Data technology to transform customer service. This system uses Spark MLlib to segment clients, provides machine learning-based predictive predictions, and handles real-time interaction management across many channels of communication. By offering customized services and predictive analytics, the platform improves the customer experience and dramatically increases sales conversions, customer satisfaction, and decreases churn. By using an inventive approach to customer service, companies can stay ahead of the curve in terms of comprehending and successfully satisfying client expectations.