Skip to content
View Vanshcloud's full-sized avatar
🌴
On vacation
🌴
On vacation
4 followers · 4 following

Block or report Vanshcloud

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Vanshcloud/README.md

Vansh Tomar

CSE Undergrad · Exploring AI, ML & Software Engineering

Still figuring out which one sticks — so I build in all of them.

LinkedIn Email

// About

  • 🎓 B.Tech CSE @ MAIT, Delhi (2023–2027)
  • 🔬 Currently deep in applied ML — gradient boosting, time-series, explainability (SHAP)
  • 🛠️ Full-stack with the MERN stack, plus Python services behind FastAPI and Docker
  • 💻 Comfortable across Python, Java, JavaScript/TypeScript, C++
  • 🧪 I care about honest evaluation — held-out seasons over random splits, tests over vibes

// Tech Stack

Languages   Python Java JavaScript TypeScript C++ SQL

Frontend   React Next.js Tailwind HTML5 CSS3

Backend & Data   Node.js Express MongoDB FastAPI DuckDB

ML / AI   scikit-learn LightGBM XGBoost TensorFlow LangChain Pandas

Tools   Git Docker Pytest Streamlit


// Featured Projects

  • ⚽ Transfer Value Predictor — Market-value prediction for professional footballers with a SHAP explanation behind every number. Evaluated on held-out seasons, not a random split. · LightGBM FastAPI Next.js

  • 🎯 Match Outcome Predictor — Calibrated home/draw/away probabilities learned from 303,517 matches across 39 competitions, scored on 62,036 walk-forward forecasts against the bookmaker's closing line. It does not beat the line, and says so. · LightGBM FastAPI Streamlit

  • 🏭 Predictive Maintenance + GenAI — Predicts equipment failure 24 hours ahead from sensor telemetry, then explains it in plain English. Dockerised, 246 tests, 86% coverage. · TensorFlow LSTM LangChain Streamlit

  • 🛡️ Intrusion Detection System — A benchmark study of flow-based attack detection on CIC-IDS2017 under leakage-aware chronological evaluation, with SHAP behind the frozen model. The same setup gives 0.025% false positives on a random split and 0.580% on a chronological one — measuring that gap is the point. · LightGBM SHAP Streamlit


I'm a Night 🦉
1  🌞 Morning       1 commits   ░░░░░░░░░░░░░░░░░░░░░░░░░     0.4%
2  🌆 Daytime     114 commits   ███████████░░░░░░░░░░░░░░    46.0%
3  🌃 Evening      83 commits   ████████░░░░░░░░░░░░░░░░░    33.5%
4  🌙 Night        50 commits   █████░░░░░░░░░░░░░░░░░░░░    20.2%

A year of contributions, popping one square at a time


What I'm listening to on Spotify


// GitHub Stats

streak

profile views

Pinned Loading

  1. Predictive-Maintenance-GenAI Predictive-Maintenance-GenAI Public

    Predict equipment failure 24 hours ahead from sensor telemetry — and explain it in plain English. TensorFlow LSTM + LangChain, served via FastAPI with a Streamlit dashboard. Dockerised, 246 tests, …

    Python

  2. Transfer-Value-Predictor Transfer-Value-Predictor Public

    Market-value prediction for professional footballers, with a SHAP explanation behind every number. Evaluated on held-out seasons, not a random split. For scouting and sports-analytics work. Python …

    Python

  3. Match-Outcome-Predictor Match-Outcome-Predictor Public

    Calibrated home/draw/away probabilities for Europe's top leagues and Brazil's Serie A, benchmarked against the bookmaker's closing line. A Streamlit dashboard shows live scores, the week's fixtures…

    Python

  4. Intrusion-Detection-System Intrusion-Detection-System Public

    Explainable ML intrusion detection on CIC-IDS2017 under leakage-aware evaluation

    Python