MD Abdur Rahim - Freelance Python and Django Developer

21/09/2025

From Python basics to full-stack web apps, here’s what I’ve leveled up with:

🎓 CS50x – Introduction to Computer Science (2022)
✅https://cs50.harvard.edu/certificates/008ada59-82c7-4f19-a99e-15c73cbd8935

🎓 CS50P – Programming with Python (2022)
✅https://cs50.harvard.edu/certificates/34ad6d3e-cd2c-4093-8319-a83e51fda36f

🎓CS50W – Web Programming with Python & JS (2023)
✅https://cs50.harvard.edu/certificates/304903fc-c244-491c-b072-4ab27394136e

🎓CS50 – Introduction to Computer Science (edX)
✅https://courses.edx.org/certificates/bcb102d30ca0411f9b60bf3b973cb788?_gl=1*b3ifrs*_gcl_au*MTMyMTUwNjcxMy4xNzU4NDc4MDk0*_ga*MTUwMjY4NjgyNi4xNzU4NDc3OTcw*_ga_D3KS4KMDT0*czE3NTg0Nzc5NzAkbzEkZzEkdDE3NTg0NzgzODUkajYwJGwwJGgw

Every course sharpened my skills & fueled my coding passion. Excited to build smarter, real-world projects!

12/08/2025

✅ Are LoRA & QLoRA still the best way to fine-tune LLMs?
👉 Short answer: Yes.
👉 Long answer: Also yes — and here’s why.
LoRA (Low-Rank Adaptation) has been the workhorse for parameter-efficient fine-tuning.
It’s fast, light on hardware, and delivers accuracy close to a full fine-tune without the big cloud bill.
Then came QLoRA — LoRA’s smarter sibling.
It keeps the same benefits but adds model quantization, meaning you can fine-tune huge models on smaller GPUs and still get great results.
For 90% of projects I’ve seen, LoRA/QLoRA gives the perfect balance of cost, speed, and accuracy.
Sure, new methods like DoRA and GaLore are on the horizon, but these two remain the most reliable choice for real-world work right now.
✅ If your business wants a custom AI solution without breaking the bank, LoRA or QLoRA is where I’d start.

Send a message to learn more

13/07/2025

Please take a look at my verified achievement from Coursera.

Credly is a global Open Badge platform that closes the gap between skills and opportunities. We work with academic institutions, corporations, and professional associations to translate learning outcomes into digital credentials that are immediately validated, managed, and shared.

13/07/2025

🚀 Exciting news! I've just developed a cutting-edge YouTube Video Summarizer using LLMs! 🎥🤖

Ever wished to grasp the essence of a lengthy YouTube video in less than a minute? Now, with my newly created Python application, it's possible! Here's what my app does:

✅ Extracts transcripts from any YouTube video
✂️ Intelligently divides the content into manageable chunks
🧠 Utilizes LangChain, Ollama, and Gemma 3B to:
🔸 Summarize the entire content
🔸 Identify key topics
🔸 Highlight impactful quotes
🤖 All this functionality is encapsulated within a LangChain agent that responds to your specific queries!

This innovation is powered locally using open-source LLMs, ensuring data security and reliability. 💻

📌 Tools Utilized:
Python, LangChain, ChatOllama, YouTubeTranscriptAPI
Model: gemma3:1b via Ollama (operates locally, no dependency on OpenAI)

🔗 Access the GitHub Repository and README here:https://github.com/abdurrahimcs50/youtube_transcript_ai_agent/blob/main/README.md

🛠 The complete code is accessible for further customization (RAG, Streamlit UI, DB integration, etc.)

If you're engaged in projects involving LLMs, LangChain, or tasks related to scraping/transcription, I welcome your feedback, suggestions, and potential collaborations! 💬

hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag

13/07/2025

Check out my latest article: How To Build a Smarter Chatbot with LangGraph https://lnkd.in/g6V3FTTM via

Learn how to build a powerful chatbot using LangGraph and LangChain — then meet the freelance Python wizard making it happen. This isn’t just a tutorial — it’s a signal to founders, CTOs, and recruiters: the next-gen backend magic is here, and it’s powered by MD Abdur Rahim.

13/07/2025

🚀 New Article Drop: How To Build a Smarter Chatbot with LangGraph
👉 https://www.linkedin.com/pulse/how-build-smarter-chatbot-langgraph-md-abdur-rahim-ncwvc/
If you're building with LLMs and struggling to manage multi-turn conversations, state, memory, or tool-chaining — this article is for you.
I break down how to use LangGraph (from the LangChain team) to build an AI chatbot with:
✅ Stateful memory
✅ Modular architecture
✅ Easy integration with OpenAI, Claude, Gemini
✅ Scalable backend logic in pure Python
This is how I build production-ready GenAI apps for my clients — with clean code, fast performance, and clear workflows.
👨‍💻 About Me
I’m MD Abdur Rahim, a Freelance Python & Django Developer. I help companies build:
🔹 AI-powered chatbots and assistants
🔹 RAG pipelines with LangChain
🔹 Scalable backends with Django, DRF, and AWS
🔹 Web scraping and data enrichment systems
📌 Portfolio: https://rahim.com.bd
Looking to build smarter with AI?
💬 Let’s connect or drop me a DM — happy to collaborate on freelance or remote projects!
hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag

Learn how to build a powerful chatbot using LangGraph and LangChain — then meet the freelance Python wizard making it happen. This isn’t just a tutorial — it’s a signal to founders, CTOs, and recruiters: the next-gen backend magic is here, and it’s powered by MD Abdur Rahim.

13/07/2025

🚀 Just Completed: Advanced Website Crawler & PDF Converter System

I've built a robust web crawler that automates document preservation by converting HTML to PDF and downloading existing PDFs - perfect for compliance, archiving, and content management!

🔍 Core Features:
✅ Intelligent recursive crawling with depth control
✅ HTML → PDF conversion with CSS/JS rendering
✅ PDF downloading with SHA-256 duplicate detection
✅ Automated scheduling (every 12 hours)
✅ Comprehensive HTML/JSON reporting
✅ Smart URL filtering (external links, admin pages)

🛠️ Tech Stack:
• Python 3.10 (Async I/O)
• aiohttp
• aiofiles
• asyncio
• beautifulsoup4
• selenium
• requests
• click
• python-decouple
• certifi
• APScheduler
• SQLAlchemy
• lxml
• html5lib
• uvloop

🌟 Key Achievements:
• Handled complex legal/financial sites: Lexis Advance & FDIC manuals.
• 87% reduction in duplicate documents.
• Configurable limits for depth/page count.
• Detailed audit trail with error tracking.

💡 Perfect for:
• Legal document preservation.
• Regulatory compliance archiving.
• Website content migration.
• Automated reporting systems.

👉 GitHub Repo:
https://github.com/abdurrahimcs50/website_crawler_pdf_converter_project

📌 Why this matters:
This project solves real-world content preservation challenges while demonstrating advanced Python skills in asynchronous operations, web scraping, and system design.

hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag hashtag

06/07/2025

🚫 Tired of your scraper getting blocked halfway through the job?
I get it. Proxy rotation, user agents, captchas... It's messy.
After building robust scraping pipelines for 20+ clients — including real estate, e-commerce, and lead gen — I boiled down my proven process into a free mini-guide:
👉 “Web Scraping Best Practices: 10 Rules to Stay Unblocked, Undetected & Efficient”
Covers real-world tricks using requests, BeautifulSoup, Selenium, anti-bot evasion, error handling, and scaling with Docker & rotating proxies.
Want it? Just comment ‘SCRAPE’ or DM me — I’ll send it over.
hashtag hashtag hashtag hashtag hashtag hashtag

06/07/2025

🎉 Just Completed: Ultimate Web Scraping for Data Science 🕸️💻
I’m excited to share that I’ve just completed the Ultimate Web Scraping for Data Science course! As a Freelance Python & Django Developer, this course has taken my web scraping skills to the next level, empowering me to extract valuable data from websites and turn it into actionable insights.
🔍 Key Skills Acquired:
BeautifulSoup & Selenium for scraping both static and dynamic websites.

Handling complex web elements like dropdowns, infinite scrolling, and alerts.

Building practical projects like a Stock Data Scraper from Yahoo Finance and Real Estate Listings extraction.

Ethical scraping practices, large-scale crawls, and data management.

Integrating scraped data into ML models or business applications.

🎓 Course Certificate:
Proud to earn a certificate for completing this course, which complements my existing experience in Backend Development, Data Analytics, and AI Integrations. You can view my certificate here: https://www.udemy.com/certificate/UC-08b96547-4b8e-46bd-ab31-db35df9f6c01/
🚀 What I Offer:
Scalable backend systems with Django, DRF, Celery, PostgreSQL.
Real-time applications using WebSockets & HTMX.
Data scraping and automation solutions using BeautifulSoup, Selenium, and Scrapy.
AI-powered applications integrating LLMs, LangChain, Langgraph, and OpenAI.

🔗 Let’s Connect:
🧑‍💻 I’ve been working as a freelancer on Upwork for over 5 years, helping clients build scalable, secure, and AI-powered applications.

🎓 Certified in CS50 (HarvardX) and Meta Back-End Development.

🛠️ Tools: Python, Django, Docker, AWS, GraphQL, REST APIs, LLMs, Generative AI, Scrapy, and more!

📣 Portfolio & GitHub:
Portfolio: https://rahim.com.bd/

GitHub: https://github.com/abdurrahimcs50

If you’re looking for a developer who can help with web scraping, backend systems, or AI solutions, feel free to reach out. I’m always open to collaborating on exciting projects!
hashtag hashtag hashtag

06/07/2025

Poll Time!
What's the hardest part of building AI-powered apps?
🔘 Setting up vector DB
🔘 Prompt engineering
🔘 Connecting LangChain
🔘 Deploying in production

Drop your vote 👇

Send a message to learn more

06/07/2025

👀 Want to see my recent work?
I’ve built AI apps, e-commerce backends, and scraping tools.

🌐 Portfolio: rahim.com.bd
💻 GitHub: github.com/abdurrahimcs50

Freelance Python & Django Developer | Data Analytics & Web Scraping Expert | Generative AI & LLM Specialist | Docker & AWS Solutions Expert | Expert In Scalable - abdurrahimcs50