Shubham Yadav

I’m an undergraduate researcher at BIOLAB where I architect and deploy machine-learning pipelines that turn complex biological data into reliable predictions. I focus on building models that are not only accurate, but also engineered for repeatability and real-world usability.

I recently worked on classifying 8,400+ protein sequences, implementing and refining Support Vector Machine and Logistic Regression models through advanced feature engineering and hyperparameter tuning. Using robust 5-fold cross-validation, I achieved 85%+ classification accuracy and iterated data-driven refinements to improve stability.

Alongside research, I’ve developed production-style backend services and APIs during my Software Development internship, supporting a centralized platform used by 200+ students and faculty. I designed secure authentication and authorization with JSON Web Tokens (JWT) and built scalable RESTful APIs for role-based access, course content management, and scheduling.

I like tackling end-to-end problems—whether it’s LLM-enhanced recommendations, real-time computer vision, or performance-focused database applications. From networking tools to model evaluation metrics like Precision@K and NDCG, I bring a systems mindset, strong fundamentals in data structures and algorithms, and a commitment to measurable outcomes.