About me
Data-driven problem solver with a passion for machine learning and a knack for creative solutions.
I find patterns in the chaos: Delivering data-driven solutions for real-world impact.
I decode data mysteries and optimize workflows with a blend of academic rigor and real-world pragmatism. From honing predictive models to fine-tuning operational efficiency, I bring tangible results to every project. From achieving near-perfect accuracy in pediatric ultrasound analysis to pioneering MLOps integration for seamless efficiency, I’ve curated a trail of innovation across fintech startups and global institutions. Whether it’s refining predictive models or spearheading data-driven transformations, I’m your go-to guru for navigating the complexities of the digital landscape. Let’s connect and chart a course for success through the realms of data mastery!
When the data screen dims, I step into the kitchen to whip up culinary delights, wield my camera to capture moments frozen in time, strum my guitar to serenade the stars, glide on skates through life’s winding paths, sketch cute doodles, and pound the pavement with my running shoes. Cooking is my flavorful escape, photography paints my world in vivid hues, guitar chords pluck at heartstrings, skating adds a dash of adventure, doodling brings whimsy to my days, and running fuels my passion for exploration. I know, it’s a lot of stuff, but hey, I love them all!
What I do
As a computer science professional with expertise in machine learning and data analysis, I can in help optimize data workflows, automate processes, and make informed, data-driven decisions.
Machine Learning
Implemented machine learning models employing various techniques to solve complex problems in areas such as natural language processing and computer vision.
Data Science
Implemented empirical methods for data analysis and statistical modeling with data visualization and interpretation.
Software Development
Implemented various projects involving software development, including web-based voice assistants, and have experience in programming languages such as Python, R, and C/C++.
Skills
My Experience
Dec 2023 - Present
Columbia University Irving Medical Center
New York, NY
Data Scientist
Achieved a stellar 98% accuracy and a 0.97 f1-score in identifying crucial anatomical structures in pediatric ultrasound videos, leveraging top-tier models like ResNet18, VGG16, and DenseNet with transfer learning, while tackling class imbalance with augmented data, and pioneering MLOps integration for streamlined lifecycle management, complemented by a user-friendly web app deployed on Render.
May 2023 - Jun 2024
Nexera.ai
Wilmington, DE
Machine Learning Engineer
Developed and scaled advanced AI infrastructure using large language models, enabling users to build, test, and automate trading strategies through plain text. Enhanced NLP model comprehension with SFT framework, boosting trade accuracy by 25%, and led R&D efforts that reduced model size by 30%, increasing deployment efficiency.
Jan 2022 - Jun 2022
JP Morgan Chase
Bengaluru, India
Data analyst intern
Boosted workforce efficiency twofold to 60% with a savvy attendance and resource system in Python and SQL, while slashing manual data tasks by 35% and ramping up accuracy with tailored Alteryx workflows. Additionally, accelerated project delivery by mentoring interns in Agile, Python, SQL, and Alteryx, trimming onboarding time by 25%.
May 2021 - Dec 2021
MoonPlexus
Pune, India
Data Scientist
Led the creation of a skin cancer detection system using TensorFlow and CNN, achieving 89% accuracy. Advanced machine learning for dermatological analysis with a multi-modal approach, increasing detection accuracy by 20%. Orchestrated cloud-based, real-time model inference on AWS, enhancing user engagement by 40% and reducing downtime.
Jan 2021 - Mar 2021
United Nations GEOLDN
Remote (South Korea)
Project data scientist
Mastered 100 TB of geospatial data, employing cutting-edge deep learning methods like CNNs to predict Land Degradation Neutrality, refining model precision by 2% with tailored architecture adjustments, and turbocharged data handling by 45% with a Python-based desktop app.