I am a Data Scientist specializing in Computer Vision, Generative AI, and machine learning solutions for industries like tobacco, oil & gas, healthcare and banking. I leverage tools like InspectAI for automated inspection and frameworks like PyTorch and TensorFlow to build innovative applications. Currently, I focus on Generative AI, developing large language model (LLM)-based systems with guardrails and context-aware protocols. Passionate about transforming data into impactful solutions, I excel at creating proofs-of-concept and production-ready AI systems.

• Languages: Python, Java
• Machine Learning: Regression, Decision Trees, SVM, Probabilistic Models (Naive Bayes, Bayesian Networks)
• Frameworks & Libraries: TensorFlow, PyTorch, Scikit-learn, OpenCV, InspectAI, LangChain, LangGraph
• Generative AI: GANs, Diffusion Models, RAG, Transformers, LLM Guardrails
• Tools & Platforms: Azure, AWS, Docker, Git, Jenkins, Postman, MongoDB, MySQL
• Other: Linux

Key projects showcasing my expertise in computer vision, Generative AI, and LLMs:

• Thought of the Day Generator
  PoC for generating daily inspirational quotes using a fine-tuned LLM.
  Technologies: Python, LangChain, Transformers

• Contract Analysis System
  PoC for automated contract analysis, extracting key clauses with NLP and LLMs.
  Technologies: Python, LangGraph, RAG

• LLM Guardrails
  PoC for enforcing safety and compliance in LLM outputs for secure AI interactions.
  Technologies: Python, LangChain, LLM Guardrails

• Service Catalog Interaction
  PoC for an LLM-based interactive service catalog with natural language queries.
  Technologies: Python, LangGraph, Transformers

• Model Context Protocol
  PoC for a context-aware protocol to improve LLM coherence in multi-turn conversations.
  Technologies: Python, LangChain, RAG

• Tobacco Leaf Quality Inspection
  Automated tobacco leaf inspection system using InspectAI and computer vision.
  Technologies: Python, InspectAI, OpenCV, TensorFlow

• Oil & Gas Pipeline Anomaly Detection
  Real-time computer vision system for detecting pipeline anomalies.
  Technologies: Python, PyTorch, OpenCV, AWS

• Healthcare Voice Q&A System
  Voice-based Q&A for healthcare using speech-to-text and RAG.
  Technologies: Python, PyTorch, RAG, Sarvam-M

Explore all my projects here.

Open to collaborations and discussions on AI and data science. Reach out via:

• LinkedIn: Mukesh Yadav
• YouTube: MukeshTecHub
• Portfolio: View my projects