Developed AI agent?powered user onboarding for next-generation learning platforms, enabling personalized and adaptive learning flows. Implemented with FastAPI, SQLModel, PostgreSQL, Docker, and deployed on Google Cloud.

• Designed and built an agentic AI system for a learning management platform that enables course creators to build, customize, and refine entire learning courses through natural-language conversation.
• Developed a multi-step Google ADK agent pipeline that interprets user intent via chat, proposes structured course modifications, visually previewed in a React-based CMS, and executes validated changes against a relational course data model through secured, human-in-the-loop tool calls.

Key capabilities delivered:

• End-to-end course generation from uploaded PDF content, guided by conversational human steering
• Intent recognition and context-aware proposal generation with real-time visual previews
• Secured tool-use layer ensuring all database mutations require explicit user approval before execution
• Seamless integration into an existing CMS (React frontend, FastAPI backend, PostgreSQL with SQLModel)
• Containerized deployment on Google Cloud via Docker