π Welcome to my profile! I'm Shubham Pandey, a passionate AI/ML enthusiast and data aficionado with a knack for turning complex challenges into innovative solutions.
π With a Master's degree in Artificial Intelligence and a proven track record in the industry, I thrive on pushing the boundaries of technology. From optimizing sales on massive datasets to crafting cutting-edge AI models, I've consistently demonstrated my ability to drive impactful results.
π My journey has taken me from validating pipelines and detecting anomalies to leading data mining efforts that have transformed search coverage and query efficiency. I've also championed automation solutions that streamline processes, reduce customer interactions, and amplify overall productivity.
π My passion for problem-solving extends beyond projects β I've contributed to impactful research, spoken at industry conferences, and mentored aspiring talents. As the founder of the AI Club at the University at Buffalo, I've created platforms for knowledge-sharing and innovation.
π‘ Beyond my technical prowess in Python, ML, DL, NLP, and CV, I'm a collaborative team player who thrives in dynamic environments. My leadership skills have led to successful collaborations and impactful outcomes, both within and outside the tech realm.
π If you're looking for an AI/ML engineer who thrives on innovation and continuous learning, or a data scientist who knows how to turn data into strategic insights, let's connect! Together, we can create solutions that shape the future.
#AI #MachineLearning #DataScience #Innovation #ProblemSolver
β’ Led and executed data mining to boost search coverage by 28%, query time by 75% employing multiclass classification
β’ Built a framework to automate detection, localization, refund of damaged items for a retail client saving ~12% per order
β’ Developed post-hoc, model agnostic explainability, debugging, debiasing product MLWorks reducing downtime by 95%
β’ Reduced customer interactions by 40% leveraging audio/text-based framework to gauge sentiment for large audio calls
β’ Boosted customer success rate by 25% by employing causal time-decay attribution model for online and retail ads
β’ Designed RNN audio spectral features based diarization framework for call summarization, improving turnaround by 40%
β’ Mentored teammates & reduced turnaround time by 15 hours weekly by validating documents & automating payments
β’ Collaborated and optimized Enterprise and Retail KYC turnaround time by 90% by verifying Identification documents
β’ Deployed an anomaly detection solution to minimize bankβs losses by 18% for real-time first party fraud detection
β’ Integrated active learning, GAN, and autoencoders for synthetic data generation, cutting ML lifecycle by 33%
Coursework: Operating Systems, Compiler Design, Analysis of Algorithm, Object-Oriented Programming
β’ Vice President at IEEE student's chapter — Built a platform to prepare ~600 students annually for competitive exams
β’ Resource Head at Sports and Adventure Club — Increased student participation by ~250%s
PCM Computer Science
CBSE, Physics/Chemistry/Math with Computer Science
β’ Head Boy & House Captain
β’ Organized CBSE ZONAL swimming competition where 1200 students from 100 schools participated for 5 days, handled data management and event schedule