Skip to content
View tarun-rai21's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Block or report tarun-rai21

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
tarun-rai21/README.md
Typing SVG
Profile Views Β  Followers Β  Focus

🧠 About Me

name     : Tarun Kumar Rai
degree   : MSc Data Science
college  : IIIT Lucknow
location : Uttar Pradesh, India

focus:
  - Machine Learning & Deep Learning
  - Large Language Models (LLMs)
  - Generative AI Systems
  - Data Structures & Algorithms

currently_building:
  - GPT-style LLM from scratch
  - Hybrid LSTM + XGBoost Forecaster
  - GenAI-powered applications

motto    : "Build. Break. Learn. Repeat."

πŸ› οΈ Tech Stack

πŸ’¬ Languages

πŸ€– ML / AI Frameworks

πŸ“Š Data Science

Pandas NumPy Matplotlib Seaborn Jupyter Kaggle

🧬 Generative AI

HuggingFace LangChain OpenAI Transformers RAG

🧰 Tools & Platforms

☁️ Data & Cloud


πŸš€ Featured Projects

🌑️ Heatwave Forecasting

Hybrid LSTM + XGBoost model for temperature prediction in Uttar Pradesh using NASA POWER meteorological data. Multi-step forecasting with residual learning architecture.

Python PyTorch XGBoost NASA API Kaggle

🧠 LLM From Scratch (WIP)

GPT-style transformer built step-by-step β€” BPE tokenizer, embeddings, multi-head attention, and full autoregressive pipeline. Guided by Raschka's book + "Attention is All You Need."

Python PyTorch Transformers BPE

πŸ“‰ Churn Risk Intelligence

End-to-end ML pipeline for telecom churn prediction with customer segmentation and Customer Lifetime Value (CLV) analysis.

Python Scikit-learn XGBoost Pandas

πŸ§ͺ Molecular Melting Point Prediction

Chemoinformatics ML pipeline using RDKit molecular descriptors to predict melting points from chemical structures.

Python RDKit Scikit-learn Chemistry

πŸ₯ Heart Attack Prediction

Classification pipeline for cardiovascular disease risk using clinical features and ensemble ML methods.

Python Scikit-learn EDA Healthcare

πŸ’» DSA & SQL Practice

Structured LeetCode DSA solutions, GFG solutions by difficulty category, and SQL interview queries for placement preparation.

C++ Python SQL Algorithms


πŸ“Š GitHub Analytics

Β 

πŸ† GitHub Trophies


πŸ“ˆ Contribution Activity


🌐 Find Me Online

LinkedIn LeetCode GeeksForGeeks Kaggle



"The best way to learn is to build something real β€” then break it, fix it, and understand why."


Pinned Loading

  1. ab-test-engine ab-test-engine Public

    A/B testing engine that enforces SRM checks, power analysis, CUPED variance reduction, and peeking-correction before reporting any effect β€” validated against synthetic data with known ground truth …

    Python

  2. aegis-gemma aegis-gemma Public

    Offline-capable AI agent combining a locally-trained LSTM heatwave forecaster with Gemma 4 agentic reasoning to deliver life-saving, location-specific survival advice β€” no internet required. Built …

    Python

  3. ecommerce-rag-engine ecommerce-rag-engine Public

    e-commerce semantic search and RAG pipeline

    Python

  4. LeetCode LeetCode Public

    This repository serves as a collection of my solutions to various LeetCode Data Structures and Algorithms (DSA) problems.

    C++

  5. TensorTonic-Solutions TensorTonic-Solutions Public

    My solutions to TensorTonic problems. TensorTonic is an interactive, browser-based education platform designed for learning machine learning (ML) and artificial intelligence (AI) by building algori…

    Python

  6. Building-LLMs-From-Scratch Building-LLMs-From-Scratch Public

    Implementing a GPT-style Large Language Model from scratch in Python & PyTorch, based on the "Attention is All You Need" paper. 🚧 Work in Progress

    Jupyter Notebook