BEGIN:VCALENDAR VERSION:2.0 PRODID:-//360exp//Events//ES METHOD:PUBLISH BEGIN:VTIMEZONE TZID:Asia/Kolkata TZURL:http://tzurl.org/zoneinfo-outlook/Asia/Kolkata X-LIC-LOCATION:Asia/Kolkata BEGIN:STANDARD TZOFFSETFROM:+0530 TZOFFSETTO:+0530 TZNAME:IST DTSTART:19700101T000000 END:STANDARD END:VTIMEZONE TIMEZONE-ID:Asia/Kolkata X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT UID:227e3684-090a-4a63-a728-06b085de1566@360exp.net SEQUENCE:0 DTSTAMP:20260911T010232 DTSTART;TZID=Asia/Kolkata:20260926T043000 DTEND;TZID=Asia/Kolkata:20260927T113000 SUMMARY:Agentic AI\, Automation and Multi-Agent Systems | 2 Days Workshop at IIT Madras Research Park LOCATION:IIT Madras Research Park DESCRIPTION:Agentic AI & Multi-Agent Systems | From Tools to Autonomous Mu lti-Agent SystemsBy Techobytes TechnologiesEligibility: Students and Worki ng professionalsOpen to all: B.Tech\, B.E.\, BCA\, MCA\, B.Sc IT\, M.Sc IT \, MBBS\, MS\, MD\, Biomedical\, Biotechnology\, BBA\, MBA\, B.A.\, and ho bbyists from all backgroundsCourse ContentDuration: 2 Days (14-16 HRS)Fee includes 2 days workshop pass with Lunch during the workshop.Go from a sin gle chatbot to a team of AI agents that plan\, act\, use tools\, and colla borate. Why this workshop?You might have already built an AI that talks. N ow you'll build an AI that acts. You'll give your models real tools\, let them query live data\, reason and self-correct\, and team them up into mul ti-agent systems that tackle tasks no single model could. Every module is hands-on - you'll watch working agents in action\, then prompt the AI to b uild and harden your own. By the end you'll have a production-style\, mult i-agent application that's truly yours.What you'll build● An AI that cal ls real tools and live APIs - and resists misuse.● An assistant that ans wers questions from a real database by writing its own SQL.● Autonomous agents that plan\, act\, and catch their own mistakes.● A team of specia lized agents (planner\, researcher\, writer\, reviewer) working together. ● An observable\, cost-aware multi-agent system as your capstone.How we' ll work - "vibe coding"● See it run first: every module opens with a liv e demo of working agents.● Then build yours: you prompt the AI in AntiGr avity to design\, extend\, and harden it.● Ship something every session - real systems\, not slideware.Before you start●You know (Generative AI Essentials)\, or equivalent - you should be comfortable building a chatbot and a basic RAG system.● Python 3.x\, Anaconda\, and AntiGravity instal led● No database to install - we'll use SQLite (built into Python) with a sample database weprovide.● Access to a foundation-model API (we'll he lp you set up keys at the start).Course modulesModule 1 - Tools & Tool-Ori ented AgentsGive your AI hands - then make those hands production-tough. ● Introduction to tool (function) calling.● Structured outputs and JSO N schema design.● Tool routing\, orchestration\, and dynamic tool select ion.● Building custom tools: calculator\, database query\, web search. ● API-based tool integration.● Handling tool failures and recovery\; t esting prompt-injection and misuse.● Demo: An LLM deciding to call a too l and using the result in its answer.● Project: Extend your Level 1 chat bot with external API tools\, then harden it.Module 2 - Talk to Your Data: SQL AgentsAsk a question in plain English - your agent writes the SQL and answers from a real database.● Why text-to-SQL: natural language as the interface to structured data.● Giving the agent a database schema as co ntext.● The SQL query tool: generate\, run\, and return results via func tion calling.● Reading results and recovering from errors - invalid quer ies\, empty results\, retries.● Guardrails: read-only access\, blocking destructive SQL\, validating output.● Demo: Ask a question in English an d watch the agent query a SQLite database live.● Project: Build a natura l-language SQL assistant over a sample database.Module 3 - From Chatbot to AgentStop just answering - start acting. Turn your chatbot into an agent that decides and does.● Transform a chatbot into an autonomous agent.● Implement perception\, planning\, and action loops.● Add memory and too l-decision logic.● Project: Build a functional autonomous AI agent.Modul e 4 - Planning & Reflection SystemsTeach your agent to think before it act s - and to catch its own mistakes.● Implement reasoning loops.● Add re flection and self-correction mechanisms.● Compare single-shot vs reflect ive outputs.● Project: Build a self-improving agent that reviews and fix es its own work.Module 5 - Multi-Agent Collaboration & Design PatternsOne agent is smart. A well-structured team of agents is unstoppable.● Define agent roles: planner\, researcher\, writer\, and reviewer.● Enable stru ctured message passing between agents.● Core patterns: simple\, sequenti al pipeline\, parallel\, critique/judge\, and hybrid.● Choose the right pattern for the task.● Demo: Design a multi-agent workflow visually and trace its execution flow.● Project: Build a working multi-agent workflow using the core patterns.Module 6 - MCP & A2A ProtocolsPlug agents into an ything - and let them talk to each other - using open standards.● Introd uction to the Model Context Protocol (MCP).● Composable AI systems.● A gent-to-Agent (A2A) communication concepts.● Context servers and modular design.● Project: Build a modular AI assistant using MCP principles.Mod ule 7 - Capstone: Your Own Multi-Agent SystemBring it all together into a real system you designed end to end.● Combine tools + a SQL/data agent + multi-agent patterns into one system.● Design a team of agents for a re al domain task.● Add planning\, reflection\, and failure recovery.● Bu ild in observability: logging\, tracing\, token/cost monitoring\, and sing le- vs multi-agent benchmarking.● Deliver a production-style\, shareable multi-agent application.By the end of Workshop\, you will be able to● B uild tool-using\, autonomous agents that plan\, act\, and self-correct.● Build agents that query real databases from natural language - safely.● Design multi-agent systems using proven agentic patterns.● Connect and compose agents using open protocols (MCP / A2A).● Make agent systems rel iable\, secure\, observable\, and cost-efficient.\n\nBoletos: https://www. district.in/events/agentic-ai-automation-multiagent-systems-2-days-worksho p-at-iit-madras-research-park-aug1-2026-buy-tickets URL;VALUE=URI:https://360exp.net/events/227e3684-090a-4a63-a728-06b085de15 66 END:VEVENT END:VCALENDAR