- Spearheaded many tasks for Digital Transformation of company’s business operations.- Performed data cleaning and standardization using Python to ensure accuracy and consistency.- Conducted data analysis using PowerBI of Booking and Sales data to derive actionable insights.- Executed ETL processes using Python, transitioning the company’s Excel datasets into MySQL DBMS.

Document-Based Data Enrichment using BoM (Bill of Materials):• Leveraged AWS Textract’s text extraction capabilities to develop an automatic data ingestion system.• Implemented a Named-Entity Recognition system using AWS Comprehend to extract entities like names, locations, organizations, and dates from unstructured text data.• Developed custom rules and fine-tuned the ML models to enhance entity recognition accuracy and handle domain-specific entities.Image Classifier ML Model:• Built an image classification model using ResNet50 in PyTorch to categorize images as equipment, nameplate, or tag, while also identifying the brand series for enhanced image processing on company portal.• Collaborated with software engineers to integrate the classification model into existing systems.Multi-Class Text Classification System:• Built a multi-class text classification solution using AWS Sagemaker.• Trained the models on labelled datasets, optimizing accuracy, precision, and recall through hyperparameter tuning.• Deployed the ML model on AWS infrastructure, ensuring scalability and real-time prediction capabilities.Text Translation System:• Built a language translation system using AWS Translate for translation of document text from different languages into English. Integrated this into existing system, enhancing data quality and accuracy.

Data Analysis of Call Center daily performance data:• Developed a customer segmentation system using sentiment analysis on call center daily performance data.• Utilized K-means clustering to segment customers based on issue topic associated with each interaction.Data Cleaning and Pre-processing Tasks:• Automated data cleaning and pre-processing tasks like tokenization, stemming, and lemmatization.• Developed a system for efficient imputation for missing data, outlier handling and text data normalization.• Conducted exploratory analysis, and employed Regex and NLP techniques to enhance text quality.WordPress Website Development for Showcasing Company Projects and Services.

CNN-Based Visual Odometry System using Deep Learning:• Developed a visual odometry system leveraging Convolutional Neural Networks (CNN).• Trained ML models to extract features from sequential image frames for precise motion estimation.• Compared CNN-based approach with traditional methods, showcasing the superiority of deep learning solution.