Skip to content

About

Hey this is Sandy

Resources

Readme
Activity

Stars

1 star

Watchers

1 watching

Forks

0 forks
Report repository
BranchesTags

Latest commit

 

History

135 Commits

Folders and files

NameName
Last commit message
Last commit date
.github
.github
 
 
assets
assets
 
 
profile-3d-contrib
profile-3d-contrib
 
 
.gitignore
.gitignore
 
 
README.md
README.md
 
 

Repository files navigation

Santosh Kumaar — Machine Learning Engineer. B.Tech CSE @ SRM · BS Computer Science @ IIT Madras · Chennai, India. Contributions, public repos, stars earned, followers, and language mix.

Featured project — AURA Preprocessor: LLM-guided data preprocessing with a human in the loop. A Groq LLM reads a raw CSV or Excel dataset and proposes cleaning strategies; you choose which fixes to apply; it outputs a cleaned dataset and suggests suitable ML models. Built with Python, Flask, pandas, scikit-learn and Groq.

Project — AI Learns To Drive: a neural network teaches itself to race with no hand-coded driving rules. Trained with PPO in PyTorch, the driver is a small Transformer reading distance sensors; re-implemented in TypeScript so the simulation and the network run client-side with no server; you can wipe the brain and watch a fresh network learn to drive live in the browser.

Project — Kira: a voice shell around a headless Claude Code session. Say a wake word or press a key, talk, and get a spoken answer; wake word, speech-to-text and the voice all run locally, with the only network call being the one Claude Code makes anyway; it drives a real Claude Code session, so it can read notes, search repositories and write things down. Built with Python, Claude Code, faster-whisper, openWakeWord and Piper.

Project — Portfolio: a scroll-driven portfolio built as its own work sample. Full-height scenes where scrolling scrubs an authored timeline; every animation gated on a reduced-motion preference with an on-page toggle; two scenes added in August so it introduces the person, not just the craft.

Tech stack — languages: Python, JavaScript, TypeScript, C++, Java, SQL. Frameworks: React, Node.js, Express, FastAPI, pandas, scikit-learn. Tools: Azure OCR, Groq API, PostgreSQL, MongoDB, Docker, Git, Vercel.

🏙️ Contribution skyline

Isometric 3D graph of my GitHub contributions over the past year, with a language breakdown and an activity radar

About

Hey this is Sandy

Resources

Readme
Activity

Stars

1 star

Watchers

1 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors