Jaykumar Kotiya

Experience

1. Ai Engineer Genai Systems, Real-time Rag & Scalable Mlops (Gpt-4, Langchain, Aws Azure)

   CitiusTech

   Jul 2024 — Present · US

   Architected LLM-powered clinical decision support systems using GPT, LLaMA, Falcon, LangChain, FAISS, Hugging Face Transformers, fine-tuning with LoRA & PEFT to deliver context-aware medical insights, improving diagnostic accuracy by 32% across multi-specialty use cases.• Deployed RAG pipelines on AWS SageMaker, Azure Machine Learning, and Kubernetes, integrating Vector Databases (Pinecone, Weaviate) for secure retrieval, enhancing information precision by 28% while maintaining HIPAA compliance.• Engineered multi-modal AI frameworks combining NLP, Computer Vision, and Time-Series Forecasting to predict patient outcomes, achieving 24% higher accuracy in chronic disease progression modeling.• Orchestrated MLOps pipelines using Kubeflow, MLFlow, TFX, Jenkins, Docker, Terraform, enabling automated CI/CD, model monitoring, and drift detection, reducing production deployment time by 40%.• Applied Explainable AI (XAI) with SHAP, LIME, Captum and interactive Plotly Dash dashboards, enhancing interpretability and ensuring 100% compliance with FDA and healthcare AI transparency guidelines.• Optimized edge AI and LLM inference models with ONNX Runtime, TensorRT, Quantization, and Pruning, lowering model size by 35% and reducing cloud compute costs by 30% while accelerating inference speed.

Education

• Gujarat Technological University (GTU)

  Bachelor of Engineering - BE, Electronics and Communication Engineering

• Northeastern University

  Master of Science - MS, Information Systems