Skip to content
View nabeelshan78's full-sized avatar
1 follower · 7 following

Highlights

  • Pro

Block or report nabeelshan78

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.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

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

Report abuse
nabeelshan78/README.md

Nabeel Shan

AI/ML Engineer & Researcher

I am a Software Engineering undergraduate at NUST ('27) specializing in building and researching robust, end-to-end intelligent systems—from implementing foundational models in NumPy to fine-tuning state-of-the-art LLMs for production.

About Me

  • 🔭 Core Focus: Specializing in advanced model adaptation (PEFT, RLHF) and performance optimization for deployment across Generative AI, Computer Vision, and Reinforcement Learning.

  • 🌱 Learning Philosophy: I believe mastery comes from building. My repositories are a testament to this, showcasing my ability to translate complex theory into practical, production-grade applications.

  • 🤝 Collaboration: I am actively seeking research collaborations and challenging open-source contributions that align with my goal of pushing the boundaries of AI. Let's connect and build something impactful.

🛠️ My Core Competencies & Tech Stack

I architect solutions with deep, first-principles knowledge of core algorithms, expert-level proficiency in modern AI frameworks, and practical experience deploying end-to-end applications.

🧠 Artificial Intelligence & Machine Learning

  • Core Expertise: Generative AI & LLMs, Computer Vision (CNNs, U-Net, YOLO), NLP (Transformers, Attention), Reinforcement Learning (RLHF, PPO)
  • Frameworks & Libraries: PyTorch, TensorFlow, Hugging Face (Transformers, PEFT), Scikit-learn, NumPy, Pandas

🚀 MLOps & Deployment

  • Docker, Git / GitHub Actions, Streamlit, Flask, Firebase, Model Optimization

💻 Languages & Core Computer Science

  • Python, C++, JavaScript / TypeScript, SQL, Algorithms & Data Structures

🌐 Web & Full-Stack Development

  • React.js, Node.js / Express.js, HTML5 / CSS3, Tailwind CSS

🎓 Education & Experience

  • 🎓 Bachelor of Engineering, Software Engineering | National University of Sciences and Technology (NUST), Islamabad (2023 - 2027)
  • 🔬 AI/ML Research Intern | CETQAP - The Centre of Excellence for Technology Quantum and Artificial Intelligence Pakistan (July 2025 - Present)
  • 💻 AI/ML Intern | Software Productivity Strategists, Inc. (SPS) (June 2025 - Present)

📊 My GitHub Activity

Nabeel's GitHub Stats Nabeel's Top Languages

Nabeel's GitHub Streak

Pinned Loading

  1. reinforcement-learning-human-feedback-scratch reinforcement-learning-human-feedback-scratch Public

    End-to-end implementation of Reinforcement Learning with Human Feedback (RLHF) to align a GPT-2 model with human preferences — covering Supervised Fine-Tuning (SFT), Reward Modeling, and PPO-based …

    Jupyter Notebook 2

  2. safe-llm-adaptation-peft-rlhf safe-llm-adaptation-peft-rlhf Public

    An end-to-end pipeline for adapting FLAN-T5 for dialogue summarization, exploring the full spectrum of modern LLM tuning. Implements and compares Full Fine-Tuning, PEFT (LoRA), and Reinforcement Le…

    Jupyter Notebook 1

  3. Transformer-Adaptation-Playbook Transformer-Adaptation-Playbook Public

    An empirical study of Transformer adaptation techniques. Pre-training from scratch (MLM), classic fine-tuning, and from-scratch implementations of PEFT methods (LoRA, Adapters). Tuning both encoder…

    Jupyter Notebook 1

  4. NLP-From-Scratch-to-Deployment NLP-From-Scratch-to-Deployment Public

    An end-to-end project where I built a Transformer from scratch, fine-tuned DistilBERT to 94.8% accuracy, and deployed it as a live interactive demo on Hugging Face.

    Jupyter Notebook 1

  5. gpt-forge-from-scratch-transformer gpt-forge-from-scratch-transformer Public

    A clean, modular implementation of a decoder-only Transformer (mini-GPT) from scratch in PyTorch for autoregressive text generation.

    Jupyter Notebook 2

  6. IntelliBot IntelliBot Public

    An advanced, end-to-end pipeline for building production-ready conversational AI agents using OpenAI's GPT-4o. This project demonstrates a complete system with security guardrails, structured outpu…

    Jupyter Notebook 1