Tech Lead (Solutions Architect + Applied AI Lead) - Digital Twin & Clinical AI Shae Group
Configured and deployed Gemini 2.5 Flash AI agent runtimes with MCP server tool augmentation. Built AI-to-AI pipelines to automate API documentation generation across 28 backend modules.
Engineered quality assurance for a production RAG pipeline using Flowise, OpenAI, and Qdrant. Designed a 66-question benchmark across 12 knowledge-base files, classified prompt injection, retrieval, and summarization failure modes, and proposed an auditor-agent architecture for clinical response safety.
Designed multi-system agentic saga orchestration with idempotent retries, compensating rollbacks, and distributed correlation tracing across Brain, Laravel, and Axiom. Integrated a multimodal AI agent using LiveKit WebRTC, Tavus, ElevenLabs, Azure OpenAI, and Qdrant retrieval across 200 knowledge-base documents.
Formalized a multi-phase agentic SDLC and built an evidence-based evaluation framework for multi-agent AI module outputs.