I'm a final-year B.Tech AI/ML student (GPA 9.6/10) who builds ML and LLM-powered applications end to end, from model choice to deployment. My projects explain their reasoning, not just output a result. I'm looking for AI engineering internships.
Turning my case studies into live, documented builds and going deeper on production ML, explainability, and reliable deployment.
For PassCheck, I first guessed at password strength with heuristics. I reached for a RandomForest to look smart, but the heuristics already worked. I only keep a model where it earns its complexity.
The loan model caught a real bug only because I looked at SHAP values, not just the score. The report explains why a loan was declined, in plain language.
Free LLM APIs rate-limit. I built a failure handler that emails me when ranking fails, so I'm never left wondering if the weekly run worked.