👾

hitesh.ml

Hey, I'm Hitesh Kumar 👋

I work in AI & Machine Learning, but at heart I'm just curious about how ideas turn into systems that people can actually use. I love tinkering with LLM agents, multi-agent setups, and AI tools that don't just sit in theory — they run, they scale, and they ship.

This space is my personal corner on the internet — part notes, part reflections, part experiments. Think of it as my digital notebook, open to anyone who wants to peek inside.

Here you'll find things like:

  • My journey — thoughts from learning, messing up, and figuring things out
  • Visions & resolutions — not rigid roadmaps, but where I see myself growing
  • Machine learning notes — beginner-friendly entries and deeper dives
  • Weekly learnings — snapshots of what I've been exploring lately

It's not polished. It's not meant to be. It's just me — learning in public, sharing what I know, and maybe sparking an idea for someone else.

Glad you're here 🙂 Take a look around.

A photo of Gion Town, Kyoto I took on March 2024.

Current Ongoings

Studying traditional ML

Learning math for ML

Mastering the language of Python

Learning to get comfy with PyTorch till 5D

Featured Projects

Multi-RAG-Agent dashboard

Multi-RAG-Agent

A system using multiple specialized AI agents to automate workflows like text summarization, research article generation, and data sanitization.

Multi-Agent SystemAdvanced RAG ImplementationWorkflow Automation
PythonStreamlitFAISSHugging FaceGemini APINext.jsFlask
LinkedIn Automate Comment dashboard

LinkedIn Automate Comment

An AI tool that scrapes LinkedIn posts, performs sentiment analysis, and automatically generates context-aware comments to boost professional engagement.

AI-Powered CommentsSentiment AnalysisAutomated Engagement
PythonSeleniumGemini APIFastAPIStreamlitPlaywright
Property Price Predictor dashboard

Property Price Predictor

An end-to-end ML pipeline for predicting real estate prices in India. It includes automated data ingestion, advanced geospatial feature engineering, and a Flask UI, achieving a 90.3% R² score.

90.3% R² ScoreGeospatial AnalysisAutomated Pipeline
PythonScikit-learnLightGBMXGBoostPandasFlask
Subway Surfer RL Agent in action

Subway Surfer RL Agent

A project using Reinforcement Learning (PPO) and computer vision to train an AI agent to automatically play the game Subway Surfers.

PPO ImplementationComputer VisionGame Automation
PythonStable-Baselines3GymnasiumOpenCVPyAutoGUI

Socials

I’m always open to chat! Please feel free to ping me at hiteshofficial0001@gmail.com or Twitter if you have any questions or just wanna talk.

    Ask me anything!
    👋