Group Discussion
This program is connected to a group. You’ll be added once you join the program.
Agentic AI Course
Build, deploy and run a production agentic AI system on Google Cloud, one lesson at a time. You start from a blank Google Cloud project and finish with DocuMind, a multi-tenant RAG and agent application that you have deployed, tested, secured and handed over. Inside are 13 modules and 90 self-paced lessons. They cover document indexing and retrieval, grounded answers with citations, evals and safety, MCP and LangGraph agents with human approval, context caching and memory, multimodal and graph RAG, Pub/Sub ingestion, self-hosted models (Ollama, vLLM), multi-agent patterns and A2A, CI/CD release and operations, and model tuning. Module 11 ends with an independent capstone and an operational handover. You need a laptop, a GitHub account and your own Google Cloud billing account, which pays for what you deploy. While the full stack is deployed it costs ₹1,057 to ₹2,666 a day at Google's Mumbai list prices, even when nobody is using it, plus per-use charges such as Gemini calls. New Google Cloud accounts get $300 of free-trial credit for 90 days. Module 0 sets a budget alert first and shows how to bring the stack down between sessions, which stops that daily cost.