Skip to content
View muneeb-amir's full-sized avatar
1 follower · 0 following

Block or report muneeb-amir

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 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
muneeb-amir/README.md

Hi there, I'm Muneeb Amir 👋

Aspiring Computer Scientist | Generative AI & Computer Vision

I am a final-year Computer Science student at FAST University, Islamabad. My core focus lies in Generative AI, Multi-Agent Systems, and Computer Vision. I have hands-on experience in full-cycle ML development, handling everything from data preprocessing to performance evaluation using PyTorch and TensorFlow.

Currently, I am expanding my engineering toolkit by studying MLOps to master the deployment, scaling, and lifecycle management of machine learning models in production environments.

🚀 What I'm Building & Learning

  • 🧠 Architectures: Working deeply with Transformers (BERT, FinBERT, GPT-40, Gemini) and Multi-Agent RAG setups.
  • 👁️ Computer Vision: Developing CNN architectures including YOLOv8, U-Net, and CycleGAN.
  • ⚙️ Specialized Techniques: Practicing Prompt Engineering (CoT, Few-shot), Multimodal Retrieval, and Semantic Segmentation.
  • 📈 Current Focus: Integrating MLOps practices to streamline and automate AI model deployment.

💻 Tech Stack

  • Languages: Python, C/C++, C#, Java, Kotlin, Assembly Language
  • ML & AI: PyTorch, TensorFlow, scikit-learn, Hugging Face, LangChain, LangGraph, OpenAI APIs, DDPG
  • Data & Tools: FAISS Vector DB, Pandas, NumPy, SQL, Docker, Git, Streamlit, Flask

🔬 Featured Projects

  • [Multi-Agent RAG Tutoring Framework] * Developed a virtual classroom utilizing GPT-40/Gemini and LangGraph to orchestrate multi-agent dialogues between teachers and students.

    • Implemented a FAISS-based RAG pipeline, evaluated via BERTScore, BLEU, and ROUGE-L.

  • [Multimodal RAG System] * Built a processing pipeline for text and visual data from financial PDFs using CLIP (image embeddings) and Sentence-BERT (text).

    • Includes a Streamlit interface enhanced with Chain-of-Thought (CoT) prompting.

  • [Waste Detection & Semantic Segmentation] * Combined YOLOv8 (object detection) and a custom U-Net (pixel-wise segmentation) on the TACO dataset.

    • Applied data augmentation (mosaic, color jitter) to optimize mAP and IoU scores.

💼 Professional Experience

  • AI/ML Research Intern @ NACTA (Dec 2025 - Mar 2026)
  • AI Engineering Intern @ RubrixCode (Aug 2025 - Nov 2025)
  • AI and App Developer @ Protect Lab (Jul 2025 - Aug 2025)
  • AI/ML Engineer Intern @ DevelopersHub Corporation (May 2025 - Jul 2025)

📫 Let's Connect

Popular repositories Loading

  1. multimodal-rag-system multimodal-rag-system Public

    Multimodal Retrieval-Augmented Generation (RAG) system integrating text + image understanding using FAISS, CLIP, SentenceTransformers, and GPT-based reasoning with a Streamlit interface.

    Python 1

  2. semantic-product-search-ranking semantic-product-search-ranking Public

    Deep learning-based semantic product search and ranking system using SBERT embeddings and MLP trained on Amazon ESCI dataset, deployed with a real-time Gradio interface.

    Python 1

  3. multi-agent-rag-tutor multi-agent-rag-tutor Public

    Multi-agent RAG-based AI tutoring system with teacher–student dialogue, FAISS retrieval, and automated evaluation.

    Python 1

  4. taco-waste-detection-segmentation taco-waste-detection-segmentation Public

    YOLOv8 and U-Net based waste detection and segmentation system using the TACO dataset for smart city applications.

    Python 1

  5. urdu-poetry-generation-comparison urdu-poetry-generation-comparison Public

    Comparative study of RNN, LSTM, and Transformer models with multiple optimizers for Urdu poetry generation, analyzing performance, efficiency, and text quality.

    Python 1

  6. muneeb-amir muneeb-amir Public