Master Generative AI - 2 Day Workshop at IIT Madras Research Park

Certified Tech Workshop Series 2026 by Techobytes with knowledge partner eDC IIT Delhi.Hands-on training, practical demonstrations, certification.Mastering Generative and Agentic AI(Exploring the Frontiers of Generative Models)DAY 1 | Foundations & Application Building1. Generative & Agentic AI EssentialsAI, ML, Deep Learning, and Generative AI; Transformer and LLM fundamentals; tokens, context windows, and hallucinations; chatbots vs. workflows vs. agents; model selection based on capability, cost, and latency.Hands-on:Explore engineering and workplace use cases, and compare model responses.2. Prompt & Context EngineeringClear instructions, few-shot examples, task decomposition, and structured outputs; supplying relevant documents and conversation context; introduction to multimodal inputs.Hands-on:Create prompts for technical explanations, coding assistance, and document extraction.3. Build Your First GenAI ApplicationCalling an LLM through a Python API; system instructions, conversation history, and JSON output validation; API key handling and basic error management.Hands-on:Build a technical assistant that explains concepts and summarizes documents.4. RAG: Chat with Your DocumentsDocument loading, chunking, embeddings, vector search, and retrieval; answers with source references; handling missing information; when to use prompting, RAG, or fine-tuning.Hands-on:Build a PDF knowledge assistant using FAISS or ChromaDB.DAY 2 | AI Agents, Automation & Project Development5. Build a Tool-Using AI AgentAgent components: model, instructions, tools, and state; function calling and tool schemas; tool selection and result handling; execution limits and failure handling.Hands-on:Build an agent that searches documents and uses a calculator or sample API.6. Agentic Workflows with LangGraphState, nodes, conditional routing, and checkpoints; conversation memory and human approval; introduction to Model Context Protocol (MCP) for connecting tools and data; overview of multi-agent patterns.Hands-on:Create a retrieve, analyze, draft, and review workflow.7. Evaluate & Improve AI SystemsTest questions, answer accuracy, source grounding, and tool-selection checks; prompt injection, data privacy, and tool permissions; trace inspection, token usage, cost, and latency.Hands-on:Test incorrect answers, missing context, and failed tool calls.8. Capstone: Agentic Knowledge AssistantCombine RAG, tool calling, and a review step; create a simple Streamlit interface; run locally and demonstrate a deployment path.Hands-on:Complete and present a working prototype for an engineering or workplace use case.What Participants Will BuildAn API-based Generative AI assistant; a document-based RAG application with source references; and a tool-using agent integrated into a practical project prototype.WORKSHOP DETAILSPractical skills for academic projects and workplace applicationsTools & TechnologiesPython • Jupyter / Google Colab • One selected LLM API • LangChain / LangGraph • FAISS or ChromaDB • StreamlitOne primary model provider and one vector store will be used throughout the practical sessions to keep the learning focused.Eligibility & RequirementsSuitable for B.E. / B.Tech students across engineering branches, BCA / MCA students, developers, IT professionals, and technical researchers.Basic Python knowledge is recommended. Prior AI experience is not required.Participants should bring a laptop. Setup instructions, starter code, and API access arrangements should be provided before the workshop.Key Benefits• Build practical AI applications for academic and workplace tasks.• Connect AI with documents, tools, and APIs.• Develop automation workflows with human review.• Understand accuracy, privacy, cost, and performance considerations.• Complete a project prototype for further portfolio development.