Open to ML internships, projects & opportunities

Making machine learning useful, not just impressive.

I'm Shanthan, a fourth-year Computer Science student at GRIET, Hyderabad. I work on machine learning and build systems that actually work.

S_

ML / CV / LLM

EST. 2022

Scroll to explore

01 / About

I am a Computer Science student specializing in Machine Learning and Deep Learning, with a focus on practical systems from data preprocessing and model training to deployment.

Currently working on deep learning projects and expanding my understanding of ML systems. I enjoy the space between research and reality: making careful implementation choices that make a model reliable in the hands of real people.

02 / Selected work

A selection of systems built around meaningful problems.

01

Two-stage pipeline · Medical imaging

Mitosis
Detector

A two-stage pipeline that takes a whole-slide pathology image and outputs precise coordinates of mitotic figures, automating a task pathologists currently perform manually. By combining YOLO and an EfficientNet-B2 classifier, it distinguishes true mitotic figures across unseen slides with different staining characteristics. Achieved F1 0.774 on 11 fully held-out test slides.

F1 0.774View source
02
82/100

LLM-powered platform

AI Fluency
Test

Evaluates how effectively people apply AI tools. It creates realistic work scenarios from an uploaded resume using GPT-4o-mini and scores responses on a 0–100 rubric for practical reasoning and AI tool-use strategy.

LLM · EvaluationVisit project
03
APR
18

AI-assisted journaling

Mind
Archive

A journaling app for writing down thoughts, reflecting on past entries, and brainstorming new ideas—with AI assistance that surfaces patterns and connections across entries.

AI · ProductivityVisit project
04

Price intelligence

Product Price
Tracker

Tracks prices across Indian e-commerce marketplaces, scrapes live prices, sends email alerts on price drops, and predicts future prices with an LSTM model.

Scraping · ML · LSTMVisit project
05

Kaggle · Regression

House Price
Prediction

A full end-to-end pipeline for the Kaggle House Prices competition: smart imputation and mixed encoding (Ordinal for high-cardinality features and OneHot for low-cardinality features), followed by XGBoost. Public leaderboard score: 14765, rank 281.

Top 300 on KaggleView source
06
shanthan.dev

Portfolio website

Portfolio
Website

A fast, minimal personal website designed to showcase my profile, highlight the projects I'm working on, and make it easy to connect.

Design · ToolingThis page
See more on GitHub

03 / Toolkit

Technologies I work with.

PythonPyTorchFastAPIscikit-learnPandasNumPyOpenCVDockerAWS EC2GitGitHubKaggleJupyter NotebookVS CodeSupabaseVercelOpenAI APIClaude APIComputer VisionModel DeploymentClaude CodeCodex

04 / Learning in public

A few formal stops along the way.

05 / Contact

Let's build something
great.

Open to ML internships, collaborative projects, and interesting opportunities.

shanthan.yxo@gmail.com