Skip to content
View Arsalan80425's full-sized avatar
4 followers · 8 following

Achievements

Achievement: Pull SharkAchievement: YOLO

Achievements

Achievement: Pull SharkAchievement: YOLO

Block or report Arsalan80425

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
Arsalan80425/README.md

👋 Hey there! I'm Mohammed Arsalan

🎓 B.Tech graduate in Data Science & Artificial Intelligence
📍 Indian Institute of Information Technology, Dharwad

I’m a Data Scientist and Generative AI Engineer with hands-on experience building production-scale AI systems. Over the past year, my focus has expanded from traditional ML and computer vision to LLMs, Retrieval-Augmented Generation (RAG), agentic AI workflows, and hybrid search systems.

I enjoy solving real-world problems by combining data, AI, and thoughtful system design, while keeping reliability, evaluation, and user experience at the center.

🚀 What I’ve Been Working On (Last Year)

🤖 Generative AI & LLM Systems

  • Built end-to-end RAG pipelines over 100K+ documents
  • Worked with LangChain, LangGraph, LangSmith, and Hugging Face
  • Designed agentic AI workflows for enterprise and maritime intelligence
  • Implemented hybrid retrieval (vector + keyword search)
  • Reduced hallucinations using source-grounded generation & confidence checks
  • Tuned chunking, reranking, and fallback retrieval paths for robustness

🔍 LLM Evaluation & Prompt Engineering

  • Built an LLM Behavioral Evaluation Suite
  • Evaluated models for:
    • Hallucination
    • Prompt injection & jailbreak resistance
    • Context poisoning
    • Reasoning consistency
  • Designed baseline vs optimized prompts and ran controlled A/B evaluations
  • Created an interactive dashboard to visualize LLM behavior differences

🧠 NLP & Applied AI

  • Health misinformation detection using SciBERT
  • Election misinformation analysis with graph-based clustering
  • Voice-based multilingual NLP system for agricultural queries
  • Semantic search over noisy real-world datasets

🧪 Previous & Ongoing AI Projects

  • 🖐️ Virtual Mouse using Hand Gestures (OpenCV + MediaPipe)
  • ⚽ Football Player Detection & Re-Identification (YOLO + ReID)
  • 🏥 Disease Prediction System (XGBoost, Random Forest, MLP – 97% precision)
  • 📊 Fraud Detection Dashboard with explainability (SHAP)
  • 🔐 Quantum Communication Simulator (BB84, QKD)

💻 Tech Stack

Programming & Data

  • Python, SQL, NumPy, Pandas, Matplotlib, Plotly

Generative AI / LLMs

  • LLMs, Prompt Engineering, RAG, Agentic AI
  • LangChain, LangGraph, LangSmith
  • Hugging Face (Transformers, Tokenizers)
  • Jina Reranker

Search & Retrieval

  • ElasticSearch (BM25, Fuzzy Search)
  • FAISS, Vector Databases
  • Semantic & Hybrid Search

ML / DL

  • Scikit-learn, PyTorch, TensorFlow

UI / Visualization

  • Power BI, Figma, Canva

Dev & Ops

  • Git, Linux, Docker, VS Code

📚 Current Interests

  • Reliable & evaluated LLM systems
  • RAG failure modes and mitigation
  • Hybrid retrieval architectures
  • Agentic AI and tool-using models
  • Scalable AutoML & MLOps platforms

♟️ Fun Fact

I’ve been selected twice for the Inter-IIIT Chess Team 🏆

🤝 Let’s Connect

I’m open to collaborating on LLMs, RAG systems, AI research, and real-world AI products.

📫 LinkedIn: https://www.linkedin.com/in/mohammed-arsalan-58543a305
🌐 Portfolio: https://arsalan-804.netlify.app

Pinned Loading

  1. Virtual-Mouse Virtual-Mouse Public

    Control your computer with the wave of a hand. This script uses your webcam to track your fingers, allowing you to move the mouse cursor and perform clicks without any physical hardware. It's a lig…

    Python

  2. AI-Powered-Health-Misinformation-Detection-Verification-System AI-Powered-Health-Misinformation-Detection-Verification-System Public

    🏥 Combating medical misinformation with AI — real-time fact-checking, evidence retrieval, and risk analysis. AI-powered health misinformation detection system using Streamlit, PubMed integration, B…

    Python 1

  3. AI-Powered-Election-Misinformation-Detection AI-Powered-Election-Misinformation-Detection Public

    🗳️ AI-Powered Election Integrity Dashboard - Real-time misinformation detection, bias analysis & sentiment tracking using HuggingFace Transformers & Streamlit

    Python

  4. Explainable-AI-Dashboard-for-Fraud-Detection Explainable-AI-Dashboard-for-Fraud-Detection Public

    Explainable AI Fraud Detection Dashboard using LightGBM, SHAP, and Streamlit – Interactive system for fraud prediction, detailed transaction analysis, SHAP explanations, counterfactual insights, an…

    Python 1

  5. Maritime-Operations-Assistant Maritime-Operations-Assistant Public

    Agentic AI system for maritime logistics — powered by LangGraph, Ollama (Mistral), and RAG for intelligent container tracking, predictive analytics, automated reporting, and communication.

    Python

  6. llm-behavioral-evaluation-suite llm-behavioral-evaluation-suite Public

    An end-to-end evaluation framework for analyzing and comparing LLM prompt behavior, instruction adherence, hallucination resistance, and reasoning quality under adversarial, conflicting, and real-w…

    JavaScript