Building LLM Applications & Production ML Systems

ML Engineer specializing in fine-tuning transformers, edge AI deployment, and RAG pipelines. Currently exploring agentic AI and vector databases.

Projects

Multi-Document RAG System

Production-ready Retrieval-Augmented Generation system enabling multi-document querying across PDFs with automatic source citations. Processes 27-page documents with sub-second response times and 90%+ answer relevance.

LangChain ChromaDB Gemini API Streamlit RAG
90%+
Answer Relevance
40%
Coherence Improvement

Sonnet Generator

Fine-tuned GPT-2 on Shakespeare's sonnets to generate real-time poetry with preserved meter and rhyme schemes. Engineered custom loss functions optimizing for poetic structure.

GPT-2 Fine-tuning PyTorch Flask Docker
90%
User Satisfaction
15%
BLEU Improvement

Real-Time Pothole Detection

Ensemble CNN (Xception + InceptionV3) for road condition monitoring. Deployed on Raspberry Pi for edge inference, demonstrating low-cost AI solution for infrastructure monitoring.

Computer Vision Transfer Learning Edge AI Raspberry Pi OpenCV
93%
Accuracy
$35
Hardware Cost

Classify-153

Wildlife species classifier processing 10K+ images across 153 classes. Optimized with ONNX runtime for real-time inference. Includes Grad-CAM visualizations for model interpretability.

InceptionV3 Xception ONNX Grad-CAM Hugging Face
96%
Accuracy
25%
Latency Reduction

Skills & Technologies

Machine Learning

  • PyTorch & TensorFlow
  • Hugging Face Transformers
  • Fine-tuning & Transfer Learning
  • Computer Vision (CNNs, Object Detection)
  • NLP & Text Generation

LLM & AI

  • LangChain & LangGraph
  • RAG Systems
  • Vector Databases (ChromaDB, Pinecone)
  • Prompt Engineering
  • Model Optimization (LoRA, QLoRA)

MLOps & Deployment

  • Docker & Containerization
  • Hugging Face Spaces
  • Git & CI/CD
  • Flask & FastAPI
  • AWS Cloud Services

Programming

  • Python (Advanced)
  • C++ (Intermediate)
  • JavaScript (Basics)
  • SQL & Data Processing
  • NumPy, Pandas, scikit-learn

Let's Connect