Subham Panda
builds AI that
ships.
"I don't just train models — I ship them."
3rd-year B.Tech AI & ML student at Centurion University (9.0 CGPA) — I build and deploy production ML systems, from OCR accessibility tools to real-time prediction engines.
A little about me
I'm a 3rd-year B.Tech CSE (AI & ML) student at Centurion University, Odisha, currently holding a 9.0 CGPA.
I've built and deployed 2 live AI applications and shipped 5 ML projects end to end — including an OCR-to-speech accessibility tool. I completed an AI/ML internship at InternPe, delivering 4 production ML models.
I'm currently going deeper into PyTorch, Transformers, and LLM fine-tuning while keeping my fundamentals sharp with DSA practice and Kaggle competitions.
Internship experience
Diabetes Classification
SVM (RBF kernel) trained on the PIMA dataset (768 samples) with StandardScaler and a train-test split, tuned to avoid overfitting.
Car Price Prediction
Feature engineering across 4,000+ records with IQR outlier removal and 8 encoded categorical fields, modeled with Gradient Boosting.
IPL Match Winner Prediction
Predicted live match outcomes using 8 in-match features — venue, teams, runs, wickets, overs, and last-5 stats — via Random Forest.
Breast Cancer Detection
Deep learning classifier (Sequential — Dense + Dropout, StandardScaler) on the Wisconsin dataset, validated with confusion matrix and classification report.
Personal projects
NeuroReader AI
An accessibility tool that extracts text from documents via file upload or a live camera feed and converts it to audio through an OCR → TTS pipeline (camera input → OpenCV preprocessing → pytesseract OCR → gTTS audio output).
Wine Quality Detection
A multi-layer ANN (Dense + Dropout) trained on 11 physicochemical features with a stratified split for generalization, wrapped in an interactive web app for instant quality predictions.
Customer Segmentation — K-Means Clustering
Clustered 200 customers using the Elbow Method to find an optimal K=5 (Silhouette Score 0.5547), identifying 5 segments — Premium, Impulsive, Careful Savers, Budget, Average — with targeted marketing recommendations.
What I work with
Academic background
B.Tech — CSE (AI & ML)
Class XII · Science (PCM)
Class X
Certifications & Activities
- Ethics in the Age of Generative AI — certification completed
- GFG Java Programming Course — in progress
- Regular LeetCode DSA practice & Kaggle competitions
- Languages: English (Professional), Hindi (Fluent), Odia (Native)
Let's talk
Looking for AI/ML internship opportunities where I can keep shipping models that actually make it to production. Based in Puri, Odisha — open to remote work.