Featured Project:
LLM-based Summarization
						Developed a LLM PDF-summarization that can be used for PDF, reducing documentation review time by 90%.
						
					
				
				Data Scientist with expertise in Machine Learning, NLP, Causal Inference, and A/B Testing. Skilled in Python, SQL, TensorFlow, PyTorch, and AWS.
I am a data scientist passionate about solving real-world problems using machine learning and data-driven strategies. With experience in NLP, A/B testing, and predictive modeling, I aim to create impactful solutions in AI-driven domains.
In the span of the last 5 years, I have been part of multiple projects in Data Science and Business Intelligence. I have experience with managing projects independently and working with stakeholders to align data-driven solutions with business goals and key metrics..
Work Experience
Projects
Organizations
Developed a LLM PDF-summarization that can be used for PDF, reducing documentation review time by 90%.
						
					
				
						Identified Spotify traffic using packet metadata, trained XGBoost model achieving 0.83 precision for general Spotify traffic and 0.98 for Spotify audio traffic.
						Record audio on an Android device using Termux, sending the audio to an AWS EC2 instance for processing, classifying the audio events using a SageMaker endpoint, and sending email alerts if certain conditions (e.g., presence of a vehicle, gunshot, or engine sound) are met.
						
					Image Classification Using YOLOv8 and Grad-CAM: A Comparative Study with ResNet50
						
					Designed and built an interactive database query system that dynamically guides users through questions for seamless and efficient data retrieval.
						
					Personal Twitch Assistant for Streamers; analyze the live-chat messsages, extract the recurring topics and provide statistics to the streamer.
						
					Automate the process of logging into an Android device running Termux via SSH, collecting sensor data (e.g., accelerometer), and posting this data to Thingsboard for visualization and analysis.
						
					Data Scientist Intern - Keck Medicine of USC (May 2024 - Present)
					- Developed LLM-based EHR summarization, reducing review time by 90%.
					- Predicted patient length of stay using Random Forest, improving resource allocation.
					- Applied SHAP values to explain ML predictions, increasing AI transparency.
Marketing Data Scientist - Proso Inc (Feb 2023 - Aug 2023)
					- Conducted A/B testing, improving conversion rates by 5%.
					- Built Tableau dashboards, uncovering $1M in upsell opportunities.
I am looking for roles in Data Science and Analytics. If you feel that my profile interests you, or you want to collaborate on any cool project with me, please feel free to reach out.