ARCHANA YADAV

Built a scalable, modular ML pipeline for anomaly detection with synthetic anomaly injection, cycle detection, and batch-wise prediction. Processed raw .dat files into structured CSVs, trained models with TensorFlow, tracked experiments and versions with MLflow, and deployed a Streamlit dashboard for real-time monitoring. Automated training, inference, and workflows for production-ready, scalable deployment.

-Enhanced a PowerPoint generator microservice with anomaly charts and tables using Python and SQL, improving data presentation. -Developed imputation techniques for anomaly detection in forecasting, boosting accuracy with Python and Postman. -Managed Python environments, Git version control, and delivered well-documented code for smooth team collaboration.

-Deployed Real Time Andon Analytics system on assembly lines for effectively tracking time losses to enhance overall line efficiency. - Developed a Digital Tracking Dashboard for Complete Knock Down Lots to optimize the packaging and shipment process while reducing delivery time and improving product quality. - Collaborated with cross-functional teams to implement industry 4.0 initiatives in five manufacturing facilities across India, driving digital transformation and enhancing operational efficiency.

-Led Quick Response Team (QRT) for BS6 vehicle complaint resolution by analyzing field data. Conducted trend-based root cause analysis and mobilized cross-functional teams via daily War Rooms, resolving 50+ critical issues. -Designed and implemented a real-time dashboard and vehicle history card by leveraging Python and Power BI for automated reporting, KPI tracking, and business metric visualization, enabling data-driven decision-making and reducing manual reporting time by 40%.

-Engaged in a project focused on identifying and validating fault codes for an electric bus. -Studied battery performance of electric buses across diverse driving conditions. -Contributed to the launch of BS6 vehicles and played a key role in prototype manufacturing.

Building machine learning based solution to identify demagnetisation related losses in permanent magnet synchronous motors using PyTorch