I'm a passionate AI developer and software engineer with a deep focus on Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG). Currently pursuing a Bachelor's degree in Software Engineering at NUML (3rd Semester), I'm driven by the challenge of building intelligent systems that solve real-world problems through clean code and best practices.

I specialize in designing production-grade AI architectures, engineering LLM pipelines, and creating meaningful open-source contributions. When I'm not coding, I'm exploring emerging AI technologies and contributing to the developer community.

• Large Language Models: Prompt Engineering, Function-Calling Pipelines, LLM Benchmarking & Evaluation
• RAG Systems: Retrieval-Augmented Generation, Semantic Search, Context Grounding
• Vector Databases & Embeddings: ChromaDB, Sentence Transformers (all-MiniLM-L6-v2), Top-K Similarity Retrieval
• LLM APIs & Frameworks: Groq API (Llama-3), Google Gemini API, HuggingFace Transformers

• Languages: Python, C++, Java, SQL (basic)
• Backend: FastAPI, REST APIs, Server-Sent Events (SSE), Async Python
• Data Science: pandas, NumPy, Matplotlib, Plotly, Seaborn, BeautifulSoup

Python | FastAPI | ChromaDB | HuggingFace | Groq/Llama-3 | Gemini

An intelligent, production-grade AI chatbot with full-stack microservices architecture designed to deliver contextually grounded conversations.

Key Achievements:

• Built a microservices architecture (FastAPI backend + React + Vite frontend) with strict separation of concerns
• Engineered a hybrid retrieval pipeline combining Top-K semantic search & keyword matching over 384-dimensional HuggingFace embeddings
• Implemented real-time token streaming via SSE with Groq (Llama-3) and Gemini 2.5 Flash function-calling
• Reduced hallucinations through production prompt engineering with context-grounding constraints (temp: 0.3) and sliding-window session memory
• Secured API with SlowAPI rate limiting, CORS middleware, and graceful fallbacks for production stability
• Conducted comprehensive LLM evaluation across Faithfulness, Answer Relevance, and Context Precision metrics

View Repository →

Python | Groq LLM | Text-to-Speech | REST APIs | BeautifulSoup

An AI-powered voice briefing assistant that fetches live data, processes it through Groq's LLM, and delivers personalized daily briefings in Urdu.

Key Highlights:

• Integrated 2 live REST APIs (Google Calendar, OpenWeatherMap) + BeautifulSoup web scraper for live news
• End-to-end Urdu text-to-speech (TTS) pipeline for voice delivery
• Modular architecture with independently deployable services for easy maintenance and upgrades
• Orchestrated multi-source data pipelines into a unified audio briefing

View Repository →

Python | pandas | Matplotlib | Seaborn | Jupyter Notebook

Comprehensive data analysis project extracting actionable insights from entertainment datasets.

Highlights:

• Cleaned & transformed multi-dimensional data (null handling, duplicate resolution, format standardization)
• Generated 8+ statistical visualizations revealing content distribution patterns across country, genre, and rating
• Translated raw data into business-readable insights

View Repository →

Python | pandas | Plotly | REST APIs | Jupyter Notebook

Interactive financial analytics dashboard for historical time-series analysis.

• Fetched & processed financial time-series data via REST APIs
• Applied pandas resampling & null-handling on historical data
• Built interactive dashboard using Plotly to visualize stock price vs. revenue divergence

View Repository →

Java | Swing/JFrame | OOP | NetBeans IDE

A full-featured dual-role GUI application demonstrating core software engineering principles.

• Role-based access control (Admin / Customer modes)
• Vehicle booking & inventory tracking system
• Applied OOP principles (inheritance, encapsulation, polymorphism) across 10+ classes

View Repository →

National University of Modern Languages (NUML)

• Coursework: Programming Fundamentals, Object-Oriented Programming, Data Structures & Algorithms, Software Requirements Engineering, Software Engineering Principles
• Activities: PR and Outreach Team Member, Google Developer Group on Campus

• Programming Methodologies (CS106A) — Stanford University / Code in Place (June 2025)
• Python for Data Science, AI & Development — IBM / Coursera (October 2025)

• Advanced LLM Architectures – Fine-tuning, RLHF, and model optimization
• Cloud Infrastructure – AWS & GCP for scalable AI systems
• System Design at Scale – Building robust, distributed systems
• Multi-Agent LLM Systems – Orchestrating complex LLM workflows

I'm always excited to collaborate on innovative AI projects, discuss LLM architectures, or explore new technologies. Feel free to reach out:

• Email: sharfmuzamil@gmail.com
• LinkedIn: linkedin.com/in/m-muzammil-

I'm an active contributor to the open-source community and a member of the Google Developer Group on Campus at NUML. I believe in the power of collaboration and shared knowledge. Check out my repositories for ongoing projects and contributions.