Passion Fuels Purpose! 

• Researched GANs, WGANs, and DCGANs through extensive literature review and paper analysis

• Presented pioneering research paper on leveraging GANs for imputing missing data, demonstrating expertise in applying novel techniques to real-world problems

• Leading the development of a full-stack client-server application utilizing FastAPI, Docker, and MongoDB

• Innovating with LLM technology, specifically Nous Hermes 2 LLM model, to create a speech-based application for recording daily finances, providing future predictions, and personalized financial advice

• Designed a React and Flask-based web app integrating OCR and NLP for named entity recognition, automating the Request for Quote to cost estimation process to provide Oil and Gas companies with quote ranges for new factory setups, achieving a 70% reduction in manual effort

• Enhanced efficacy of custom Mask-RCNN model involving extraction of lines and diagrams from P&ID by 15%

• Deployed Object Detection model on Raspberry Pi to automate casting defects detection, achieving 78% accuracy. Created an monitoring website, which tracks castings, marks exact problems with bounding boxes, and generates detailed PDF reports identifying specific issues

• Tailored a custom Optical Character Recognition Architecture using neural networks with 93% extraction accuracy

• Implemented Bayesian methods to develop a probabilistic model for predicting customer churn, improving retention strategies and increasing customer lifetime value by 20%

• Employed Decision Tree, K-means Clustering, and Apriori algorithms to discover underperforming ads

• Implemented strategic adjustments resulting in a remarkable 15% increase in overall campaign effectiveness

• Applied time-series analysis to optimize ad performance, leading to substantial cost savings of over $10,000

• Utilized Heatmaps and Geospatial Visualization techniques to present findings, enhancing data comprehension

• Developed dashboards in Power BI to monitor key performance indicators (KPIs) and providing actionable insights

• Implemented Google's Neural Style Transfer research paper, showcasing the fusion of artistic styles with images

• Researched advanced image classification techniques capable of distinguishing between over 1000 classes

• Authored findings from in-depth research in Neural Machine Translation and Deep Local Features on Medium`

• Developed an indoor navigation app utilizing A* algorithm to curtail walking time on university campus

• Leveraged Flutter to develop a user-friendly application helping new visitors navigate the campus with ease