🤖 Generative AI & Agentic Document Intelligence Architected an end-to-end Generative AI pipeline using transformer-based LLMs, enabling intelligent document categorization and context-aware semantic search across enterprise repositories Engineered agentic AI workflows with multi-step reasoning chains, implementing MCP (Model Context Protocol) and A2A (Agent-to-Agent) interaction patterns for autonomous document processing Deployed a hybrid RAG architecture combining FAISS vector similarity search with structured queries, improving LLM grounding and contextual retrieval accuracy by reducing irrelevant results Containerized and deployed scalable model inference pipelines using Docker on Azure Kubernetes Service (AKS), ensuring production-grade availability and horizontal scalability 🧠 LLM Fine-tuning & NLP Classification Fine-tuned BERT and GPT-family models for multi-class ticket classification, automating customer support routing across thousands of daily conversations Built automated CI/CD workflows for model training and deployment using Kubeflow Pipelines, cutting manual deployment effort and enabling continuous model improvement Optimized model inference for production SLAs and integrated via REST APIs into live customer-facing microservices, reducing response latency 📊 Predictive AI & Cloud Deployment (GCP) Designed and deployed customer purchase prediction models on Google Cloud AI Platform, enabling real-time inference for high-throughput transactional workloads Engineered features from large-scale MySQL datasets using Python (Pandas, NumPy), applying statistical modeling and EDA to surface key behavioral signals driving purchase decisions