SAGAR BHAGWATKAR

With over 2.5 years of comprehensive experience as a data scientist and machine learning engineer, augmented by a Masters in Data Science and Innovation from the University of Technology Sydney with top grades, I possess a proven track record of leveraging various tools and techniques to empower businesses to make informed decisions based on data. I am enthusiastic about applying my seasoned expertise to spearhead innovative projects and catalyze transformative decision-making processes.

• Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization – COURSERA
• Structuring Machine Learning Projects
• Neutral Network and Deep Learning
• Machine Learning with Python
• Data Analysis with Python
• Data Visualization with Python
• Data Science Methodology
• Databases and SQL for Data Science

ABN Lookup Tool, Centre for WHS, NSW

· Utilized Python for web scraping tasks on the Yellow Pages website and the ABN lookup website, extracting pertinent information such as business name, location, pin code, and ABN number.

· Employed a differential analysis approach to identify discrepancies in ABN numbers between the Yellow Pages website and the ABN lookup website. Businesses with disparate ABN numbers underwent further verification by querying their respective websites to ascertain the accurate ABN number, if available.

· Implemented Python scripts to parse relevant keywords from the Australian Business Register (ABR) website, specifically focusing on terms related to fall incidents.

· Developed a user interface leveraging Streamlit framework. Users input a specific business activity name, prompting the application to generate a list of businesses susceptible to the indicated fall incident.

· Incorporated additional search refinement functionalities within the interface, allowing users to specify parameters such as pin code, location, and desired number of search results.

· Integrated comprehensive documentation within the application interface, offering users guidance on optimal utilization of the tool and enhancing their proficiency in navigating its functionalities.

Wafer Fault Detection, Applied Materials

· Implemented a cutting-edge wafer fault detection system utilizing the YOLO (You Only Look Once) object detection framework.

· Curated a comprehensive dataset comprising annotated wafer images delineating various fault types with corresponding bounding boxes.

· Preprocessed the dataset to standardize image dimensions, normalize pixel values, and convert annotations into the YOLO format for seamless integration with the model.

· Configured the YOLOv3 model by leveraging pre-trained weights and fine-tuning the network architecture to accommodate the nuances of wafer fault detection.

· Orchestrated the training process, optimizing model performance through iterative adjustments of hyperparameters and augmentation techniques.

· Evaluated the trained YOLO model on a dedicated validation set, meticulously assessing its precision, recall, and mean average precision (mAP) to ensure robust fault detection capabilities.

· Leveraged the deployed YOLO model to conduct real-time inference on new wafer images, accurately detecting faults and visualizing bounding boxes around identified anomalies.

· Collaborated closely with domain experts and production engineers to validate the efficacy of the system in real-world manufacturing environments, driving significant improvements in defect detection rates and operational efficiency.

• DATA HEAD, PHOENIX 2018, YCCE
• ORGANIZING COMMITTEE MEMBER, PRATIKRUTI 2018, YCCE
• CO-HEAD TECHNICAL COMMITTEE, MECHFIESTA 17.0, YCCE
• EVENT MANAGEMENT MEMBER, YASH 17.0, YCCE
• COMMITTEE MEMBER, MECHFIESTA 16.0, YCCE