Develop AI/LLM components — prompt engineering with schema-constrained output, LLM API integration (Anthropic/OpenAI/Azure OpenAI), and RAG or classification pipelines that map legacy actions to modern equivalents. Integrate the solution with the surrounding ecosystem via REST APIs, webhooks, Azure CLI/Azure AD authentication flows, and MCP server endpoints. Own quality through automated workflow validation tests and eval harnesses that score model output fidelity, and document unsupported or manually reviewed conversion cases. Build and maintain the parsing/conversion engine (Node.js/TypeScript or Python) that reads XML/JSON-based Nintex workflow definitions and generates equivalent Power Automate flow logic.   AI Developing Skill: Core Programming & Parsing Node.js / TypeScript or Python (framework matching the Agent's stack) XML/JSON parsing experience (Nintex NWF is XML-based) Experience building parsers, ASTs, or rule engines Regex / expression-language conversion experience AI / LLM Engineering Prompt engineering / structured (schema-constrained) output generation LLM API integration (Anthropic, OpenAI, Azure OpenAI) RAG or classification pipeline experience Coding Agent experience via Claude / Github copilot Eval-harness / model output scoring experience Integration / Ecosystem MCP (Model Context Protocol) server development Azure CLI / Azure AD authentication flows REST API integration & webhook handling Testing & Quality Test automation for workflow/flow validation Application level testing best practices Nice to Have (General) SharePoint / M365 admin experience