NuhmanPK.

Senior Software Engineer (Full Stack & AI/ML)

Building open-source tools and scalable systems used in production.

PyPI downloads
150k+
PyPI packages
15+
VS Code installs
5.3k+
VS Code extensions
3+
npm downloads
1.3k+
npm packages
5+
GitHub stars
900+
Public repos
100+
Certifications
20+
Years shipping
7+
01 / Intro

About

I’m Nuhman, a Senior Software Engineer and open-source builder with years of hands-on experience starting early in my career. I focus on creating practical, scalable software across full-stack development, automation, and AI/ML, with an emphasis on tools that solve real problems and are used in production.

01 / The journey

My journey started with Python, automation and Telegram bots: small tools built to solve problems I actually had. That pulled me deep into web scraping, crawling and browser automation: HTTP, APIs, parsing, and turning messy web data into clean, structured information. From there I grew into full-stack engineering with React, Angular, TypeScript, Node.js, Express, FastAPI, MongoDB and SQL, so I can own a product end to end, from data pipeline and API to the UI people use.

02 / AI & machine learning

Today most of my curiosity goes into AI/ML. I build practical AI systems rather than thin API wrappers: collecting and cleaning data, generating instruction datasets, fine-tuning and evaluating models, and building RAG, embeddings, semantic search and agent workflows that plug into real backends. Before LLMs I spent a lot of time on computer vision: OpenCV, OCR, object detection and automatic number-plate recognition .

03 / Open source

Open source is how I learn in public. I’ve published 8 PyPI packages, 3 VS Code extensions and 5 npm packages, plus models and datasets on Hugging Face. YoutubeTags alone has passed 100,000 downloads, and my projects have collected 900+ GitHub stars. I’m a Hacktoberfest Supercontributor, hold 16 certifications across LangChain, LangGraph, MCP, PyTorch and Google Cloud, and write hands-on guides on Medium.

04 / How I work

I’m implementation-first: I’d rather build a working prototype than debate one. I prefer automation over repetitive work, reusable tools over one-off scripts, clean and maintainable code, and measuring whether something actually works. I pick technology for the problem rather than the hype, while still experimenting with whatever is new in LLMs and developer tooling.

05 / Off-screen

Away from the keyboard you’ll usually find me hiking, brewing coffee, or exploring somewhere new. That same curiosity is what keeps me building side projects: I like understanding how things work, and then making them work better.

06 / What’s next

I’m increasingly interested in the space where AI meets real software: taking existing products, workflows and messy systems and figuring out where intelligence, automation or better tooling can make them genuinely more useful. I want to keep building open-source tools, experimenting with new models and developer workflows, and working on problems where the interesting part isn’t just making something work, but making it worth using.

07 / TL;DR

Python → automation → scraping → full-stack → AI/ML → open source. I build things, break things, learn how they work, and usually automate them so I don't have to do them twice. Most importantly, I like turning interesting problems into useful software.

02 / Career

Experience

Education

  • 2019 to 2022Bachelor's Degree in Computer ScienceUniversity of Calicut
03 / Stack

Toolbox

01JavaScript02TypeScript03React04Node.js05Express06Angular07Python08FastAPI09Telegram Bot Frameworks10Web Scraping11Browser Automation12Puppeteer13Selenium14MongoDB15SQL16Machine Learning17Computervision18GenAI Development19Docker20Aws21Pytorch22Langchain23Rag24Agents
04 / Selected work

Projects

02

Headless Driver

Headless Chrome WebDriver wrapper for Python (Selenium made simple)

Open SourcePypi PackagePythonHeadlessAutomation
03

Super Logger

Super Logger is a versatile Visual Studio Code extension that enhances your coding experience across various programming languages. It provides entertaining and informative print statements with random emojis and jokes, making debugging a delightful and effective process.

Side ProjectOpen SourceVsCode ExtensionTypescriptNode
04

Telegram Bots

Unlock a world of automated efficiency with my wide range of Telegram bots, expertly crafted for specific tasks and built to seamlessly scale

Side ProjectPythonPyrogramMongoDBNodeTypescript
05

generateApiKey

Unique api key generation tools with more built in functionalities like time / date based hashing, prefix for more unique key creation.

PythonSide ProjectOpen SourcePypi Package
06

Webtrench

A powerful and easy-to-use web scrapper for collecting data from the web. Supports scraping of images, text, videos, meta data, and more. Ideal for machine learning and deep learning engineers. Download and extract data with just one line of code

PythonSide ProjectOpen SourcePypi Package
07

SnapNutriBot

Diet tracker Bot built for google competition, powered by Google Gemini API.

Open SourcePypi PackagePythonTelegram Bot
08

Quick Llama

A Python wrapper for Ollama that simplifies managing and interacting with LLMs.

Open SourcePypi PackagePythonOllamaLLM
05 / Publications

Writing

06 / Always learning

Certifications

07 / Community

Badges