Expert AI, Software Engineer and Architect who translates complex enterprise workflows into production-grade, multi-agent AI platforms. Fuses hands-on deep learning with robust C#/.NET and Python software engineering to deliver secure, cost-optimized, and highly observable cloud-native ML systems across fintech, healthcare, and industrial automation. - Agentic AI & RAG: Architected production-grade multi-agent platforms and agentic RAG workflows using LangGraph, AutoGen, and Model Context Protocol (MCP). Built multi-stage, hybrid keyword-vector retrieval systems using Qdrant, pgvector, and custom cross-encoder reranker models. - LLMOps & Infrastructure: Established end-to-end LLMOps pipelines across Azure (Azure OpenAI), AWS (Bedrock, SageMaker), and GCP using vLLM and Triton. Directed distributed model fine-tuning (LoRA, QLoRA) and quantization strategies to minimize latency and cloud compute spend. - Intelligent Automation & Vision: Engineered high-throughput Multimodal Vision and Intelligent Document Processing (IDP) pipelines utilizing advanced OCR, layout analysis, and constrained JSON extraction integrated directly into ERP and core banking APIs. - Governance, Safety & Eval: Implemented rigorous LLM evaluation frameworks (Ragas, TruLens, HealthBench) to monitor faithfulness and hallucination risks. Enforced enterprise-grade safety guardrails, Entra ID-backed security, and Human-in-the-Loop (HITL) audit trails. - Technical Leadership: Led cross-functional engineering teams and applied research labs. Shipped specialized domain-specific models that topped benchmarking lists and ranked among Australia’s top-downloaded open-source assets. Technical Stack: - AI/ML/LLMOps: PyTorch, Hugging Face, LangChain, LlamaIndex, vLLM, Triton, MLflow, Weights & Biases. - Backend & Data: .NET, FastAPI, Springboot REST/gRPC, Docker, Kubernetes, Terraform, MSSQL, PostgreSQL, Apache Kafka. _Website not available. Sign in: https://www.twine.net/signup_…Expert AI, Software Engineer and Architect who translates complex enterprise workflows into production-grade, multi-agent AI platforms. Fuses hands-on deep learning with robust C#/.NET and Python software engineering to deliver secure, cost-optimized, and highly observable cloud-native ML systems across fintech, healthcare, and industrial automation. - Agentic AI & RAG: Architected production-grade multi-agent platforms and agentic RAG workflows using LangGraph, AutoGen, and Model Context Protocol (MCP). Built multi-stage, hybrid keyword-vector retrieval systems using Qdrant, pgvector, and custom cross-encoder reranker models. - LLMOps & Infrastructure: Established end-to-end LLMOps pipelines across Azure (Azure OpenAI), AWS (Bedrock, SageMaker), and GCP using vLLM and Triton. Directed distributed model fine-tuning (LoRA, QLoRA) and quantization strategies to minimize latency and cloud compute spend. - Intelligent Automation & Vision: Engineered high-throughput Multimodal Vision and Intelligent Document Processing (IDP) pipelines utilizing advanced OCR, layout analysis, and constrained JSON extraction integrated directly into ERP and core banking APIs. - Governance, Safety & Eval: Implemented rigorous LLM evaluation frameworks (Ragas, TruLens, HealthBench) to monitor faithfulness and hallucination risks. Enforced enterprise-grade safety guardrails, Entra ID-backed security, and Human-in-the-Loop (HITL) audit trails. - Technical Leadership: Led cross-functional engineering teams and applied research labs. Shipped specialized domain-specific models that topped benchmarking lists and ranked among Australia’s top-downloaded open-source assets. Technical Stack: - AI/ML/LLMOps: PyTorch, Hugging Face, LangChain, LlamaIndex, vLLM, Triton, MLflow, Weights & Biases. - Backend & Data: .NET, FastAPI, Springboot REST/gRPC, Docker, Kubernetes, Terraform, MSSQL, PostgreSQL, Apache Kafka. _Website not available. Sign in: https://www.twine.net/signup_WWWW…