Local-first data tool
SignalDesk
A CSV analysis workspace with read-only SQL, controlled pandas operations, charts, and evidence linked to each answer.
Software engineering · Applied machine learning
I build useful software and learn by shipping it.
I’m Shlok Bhutani, a computer science student at UNC Chapel Hill. My work spans full-stack products, practical ML experiments, and production websites.
01
Selected work
Products first, experiments second. Each project links to its source or live site.
Computer vision study
Retinopathy Grading
An ordinal PyTorch model for five-grade diabetic-retinopathy classification, with calibration, cross-dataset evaluation, and Grad-CAM.
Full-stack product
GlobeReady
An international-student workspace with authentication, document storage, reminders, and document explanations backed by secure server APIs.
04
RK Sanitarywares
A production retail website serving 500+ sanitaryware products through a custom domain for a Surat business.
Next.js · React · Custom domain
↗05
NudgeFund
A cross-platform behavioral-finance app with Supabase authentication, row-level security, and a demo mode.
Expo · React Native · Supabase
↗06
Churn Probability
A reproducible churn-risk demo with holdout evaluation, threshold selection, and a Streamlit interface.
scikit-learn · Streamlit
↗07
ML Playground
Six machine-learning algorithms implemented with NumPy and tested against their mathematical behavior.
Python · NumPy · Pytest
↗08
DOGE Forecast Lab
A time-ordered forecasting study with quantile models, baseline comparison, and honest holdout reporting.
Time series · Quantile models
↗09
AI Outlook
An Outlook add-in that reads the open email and drafts an editable, tone-aware reply.
Office.js · Express · OpenRouter
↗10
Tword
A two-word deduction game with Firebase synchronization and tested Wordle-style scoring.
JavaScript · Firebase
↗11
Price Tracker
A FastAPI price-tracking demo with safe URL validation, metadata parsing, and SQLite history.
FastAPI · SQLite · Testing
↗02
About
I’m studying computer science at UNC Chapel Hill and am interested in software engineering, machine learning, and products that make complicated workflows easier.
I like working across the full path from an initial problem to an interface people can use: data modeling, APIs, evaluation, product decisions, and the details that make a project dependable.
- Based in
- Chapel Hill, North Carolina
- Focus
- Software engineering and applied ML
- Currently
- Open to internships and project work
03
Let’s talk about useful work.
The fastest way to reach me is by email. You can also review the code behind these projects or connect with me on LinkedIn.