Module 1 – Introduction to Agentic AI (8 hours)
- What is Agentic AI? Difference between LLMs and AI agents
- Core concepts: autonomy, memory, tool use, environment interaction
- 2025 trends: multi-modal agents, AI co-workers, decentralized AI agents
- Popular frameworks: LangChain, CrewAI, AutoGen, OpenAI Assistants API
- Agentic AI use cases: research automation, sales assistants, workflow orchestration
Hands-on:
- • Build your first simple agent with LangChain
- • Try an AutoGen template for a research bot
Module 2 – Architecture of AI Agents (10 hours)
- Agent loop: Planning → Execution → Observation → Reflection
- Types of agents: reactive, planning, goal-based, collaborative
- Memory in agents: short-term vs long-term memory
- Tool integration: APIs, databases, file systems
- Agent safety & guardrails
Hands-on:
- • Create an agent that fetches real-time data from APIs
- • Add memory to an agent using a vector database (Pinecone, FAISS)
Module 3 – Building Single Agents (12 hours)
- LangChain agents: Tools, Chains, Executors
- OpenAI function calling & Assistants API
- Prompt engineering for agents
- Adding multi-modal capabilities (text, image, audio inputs)
- Latest trend: agents using Retrieval-Augmented Generation (RAG)
Hands-on:
- • Create an agent that answers business questions from PDFs
- • Add image analysis to an agent using OpenAI Vision API
Module 4 – Multi-Agent Systems (12 hours)
- Multi-agent collaboration patterns: parallel, sequential, hierarchical
- CrewAI, AutoGen for agent collaboration
- Assigning specialized roles to agents
- Orchestrating agents for large tasks (project management, research pipelines)
- Latest trend: autonomous AI teams for enterprise workflows
Hands-on:
- • Create a research team with 3 agents (researcher, summarizer, presenter)
- • Run a multi-agent project to compile & format a market report
Module 5 – Deployment & Integration (10 hours)
- Connecting agents to business tools (Slack, Notion, CRM, Google Workspace)
- Deploying agents to the cloud (Streamlit, FastAPI, Hugging Face Spaces)
- Scheduling & monitoring agent workflows
- Security, ethics, and responsible agent deployment
Hands-on:
- • Deploy a customer support AI agent to a web app
- • Set up a daily report generator with autonomous scheduling
Module 6 – Capstone Project (20 hours)
Project Theme:
- "Autonomous AI Solution for a Real-World Task" – Learner’s design and deploy a multi-agent system:
- 1. Identify a business or personal use case
- 2. Design agent roles and communication flow
- 3. Implement tool integrations & memory
- 4. Deploy to a cloud platform with a user interface
Examples:
- • Automated market research & report generation team
- • Personal productivity assistant (scheduling, summarizing, task tracking)
- • Customer onboarding AI (forms, emails, Q&A)
Final Deliverables:
- • Working agent system (GitHub code + demo)
- • Project documentation & flow diagram
- • Recorded video presentation
Assessment & Certification
- • Weekly Quizzes: 20%
- • Assignments: 30%
- • Capstone Project: 50%
- • Minimum 60% score for certification
Reviews
There are no reviews yet.