Deepak Kumar Mishra User

I’m an AI/ML-focused engineering student who turns data into intelligent, deployable systems. I build end-to-end ML pipelines, with a strong emphasis on deep learning and computer vision that solve real-world problems.

In my Research Intern role at IIT Mandi’s ACS Lab, I worked on EEG-based cognitive workload analysis using the N-Back Task. I collected, processed, and analyzed EEG data, then performed workload assessment and statistical analysis using NASA TLX datasets to understand workload variations across experimental conditions.

On the product side, I developed an AI Attendance Management System using 2 recognition pipelines (Face + Voice). I implemented FaceID login with 128-dimensional facial embeddings, dlib models, and SVM classification, and built role-based Teacher/Student dashboards with secure Streamlit UI, bcrypt authentication, and Supabase-backed data across 5+ tables.

I also enjoy building ML models that are both accurate and usable. I created a Neural Style Transfer system with VGG19 and multiple losses (content, style, and total variation), and I built CreditWise, an end-to-end loan approval prediction pipeline with EDA, scikit-learn modeling, Joblib export, and an interactive Streamlit eligibility interface.