Jason M.

AI/ML Engineer - GenAI, Agents, Traditional ML, All CSPs, Healthcare

🚀 🥇 Expert-Vetted | Hands-On AI/ML Engineer | I Build LLM Apps, RAG Systems & AI Agents (MCP, LangGraph) | Python, AWS, GCP, Azure | Healthcare & FinTech 👁‍🗨 Overview I build and ship production AI systems myself, end to end. No handoffs, no delegation: I design the architecture, write the code, and stay on it until it is deployed, monitored, and generating ROI. I bring 15+ years of hands-on AI/ML engineering, a PhD in Machine Learning from Iowa State University, and a Master's in Computational Neuroscience from UC San Diego. ✅ What I Build: • LLM Applications: RAG pipelines, chatbots and copilots, document AI, semantic search, structured data extraction • AI Agents: Multi-agent systems, MCP (Model Context Protocol) tool integrations, LangGraph orchestration, function/tool calling, agentic workflow automation • Model Customization: Fine-tuning (LoRA/QLoRA, RLHF/DPO), prompt optimization, evals and guardrails, open-weight model serving (vLLM) • Healthcare AI: Clinical trial automation, medical document generation, HIPAA-compliant systems • Full-Stack AI Products: Python/FastAPI backends, React frontends, Kubernetes, CI/CD across AWS, GCP, Azure 🎯 Recent Hands-On Builds: • Engineered a clinical trial intelligence system for enterprise pharma: ingested, embedded, and indexed 100K+ trials with multi-index, multi-LLM RAG and advanced PDF parsing, powering Q&A, chat, and benchmarking • Built a GenAI product that drafts 100+ page regulatory clinical trial protocols (95% of the full M11 document), with multi-agent validation, consistency, and styling checks • Coded and deployed an ICD-10 billing code prediction model on GCP and an EHR-integrated physician sidebar on AWS EKS • Rescued a failing third-party ML platform, refactored it, and took it to production on AWS at ResMed (NYSE: RMD), enabling their first commercial AI healthcare product • Built ML-powered ad targeting and recommendation systems generating $100K+/month, plus AI products earning $1M+ revenue in year one 💼 Industry Expertise: • Healthcare/Pharma: Clinical trials, EHR API integration, medical AI, FDA-regulated software • FinTech: Real-time fraud detection, card-linked platforms, transactional APIs (MasterCard and Visa partnerships) • Enterprise SaaS and Retail/E-commerce: Multi-tenant APIs, recommendation engines, customer analytics 🔧 Technical Stack: AI/ML: GPT-5, Claude, Gemini, Llama, DeepSeek, Qwen; fine-tuning (LoRA/QLoRA, PEFT, RLHF/DPO); RAG and GraphRAG, hybrid search, rerankers, embeddings Agents: MCP, LangGraph, LangChain, LlamaIndex, CrewAI, OpenAI Agents SDK, structured outputs, tool calling Languages: Python, TypeScript/JavaScript, SQL, Java, Go Frameworks: PyTorch, Hugging Face, FastAPI, React, TensorFlow, Scikit-learn Serving & MLOps: vLLM, Ollama, AWS (SageMaker, Lambda, ECS/EKS), GCP (Vertex AI), Azure AI Foundry, Kubernetes, Docker, MLflow, Weights & Biases Evals & Observability: LangSmith, Langfuse, RAGAS, guardrails, LLM cost optimization Databases: PostgreSQL/pgvector, Pinecone, Qdrant, Weaviate, ChromaDB, Elasticsearch, MongoDB, Redis 📊 Quantifiable Impact: • 100K+ clinical trials processed, indexed, and made queryable for enterprise users • 100+ page medical documents generated with regulatory compliance • 94% accuracy in crisis detection and 73% engagement increase for a nonprofit youth chatbot • 10X subscriber growth driven by models I built and deployed • $100K+/month revenue from ML-powered ad targeting 🎓 Credentials: • PhD, Machine Learning (Iowa State University); M.Sci., Computational Neuroscience (UC San Diego) • IBM Certified: RAG and Agentic AI; Deep Learning Specialization (Coursera) • Published AI/ML researcher (Psychological Science, ICSE); 4 provisional patents in AI/computer vision 🌟 What Sets Me Apart: I am senior, and I still write the code. On every engagement you get one engineer doing the actual work: architecting, coding, testing, deploying, documenting. Because I have built AI in regulated healthcare and fintech environments, compliance, evals, and monitoring are baked in from day one rather than bolted on. My neuroscience background shapes how I build AI systems that genuinely understand human behavior and needs. 🤝 Working With Me: You work directly with me, and I personally do the work. Expect working code early (usually in the first week), frequent demos, clear async communication, and clean documentation at handover. US-based in San Diego (Pacific time), available for both short sprints and long-term builds. Have an AI feature or product that needs to get built? Send me the details and I will reply with exactly how I would build it.