Techfest Agentic AI Workshop

Agentic AI Workshop: Understand the shift from reactive chatbots to autonomous, tool-using AI systems that sense, think, act, reflect, and operate with standard levels of autonomy while mitigating common failure modes Master function calling, token optimization, JSON schema outputs, system prompt engineering, and model selection trade-offs across speed, cost, and accuracy Implement advanced execution and reasoning patterns such as ReAct, Reflexion, Tree-of-Thought, task decomposition, and Orchestrator–Worker frameworks Connect AI models to external tools and data sources using the Model Context Protocol (MCP) standard across stdio and HTTP/SSE transports Build multi-agent collaboration systems using peer-to-peer and hierarchical topologies with frameworks like LangGraph, CrewAI, and AutoGen Architect short-term dynamic context and long-term persistent memory using vector databases, RAG pipelines, and state management Design and deploy a production-grade capstone agent featuring human-in-the-loop approvals, safety guardrails, input sanitization, cost benchmarking, and multi-platform deployment Workshop Content: Basics of Agentic AI LLM to Agent Bridge Agent Architecture & Reasoning Patterns Tool Use & Function Calling Model Context Protocol (MCP) Memory, Context & RAG Multi-Agent Systems & Orchestration Capstone: Build, Evaluate & Deploy