● Designed and developed a computer vision system to analyze palm lines and facial features for personality insights using Roboflow for annotation and YOLOv8 for detection. ● Built an end-to-end ETL pipeline to extract, clean, and transform company data for injection into OpenAI based chatbot agents, enabling enriched knowledge retrieval from internal databases. ● Developed machine learning and deep learning models (Sarima, Holt-Winters, Linear Regression, XGBoost) for accurate forecasting of global and regional product demand and sales. ● Engineered AI agents with OpenAI SDK to perform intelligent tasks including database querying, document search, web scraping, and image understanding; deployed via Vercel using a customized chatbot UI. ● Scraped government and third-party tender portals using Selenium; used LLMs to classify tender-product relevance and integrated the process into GitLab CI/CD for continuous intelligence delivery.

● Annotating 5000 images from FLIR cameras with 10 object classes using LabelMe into COCO dataset format in 2 weeks. ● Conducting object detection distance estimation research using YOLOv5 pre-trained model with a custom dataset. ● Conducting research on developing a novel single-image rain removal model using CapsuleGAN. ● Collecting over 12000 pairs of rain and non-rain dataset.