Login
Course Overview:
The Agentic AI Course is designed to provide an in-depth understanding of autonomous AI agents, their architecture, and real-world applications. Participants will learn how to build AI agents that can autonomously make decisions, execute tasks, and adapt to dynamic environments. The course covers multi-agent systems, reinforcement learning, cognitive AI, LLM-powered agents, and real-world deployment strategies.
Day 1 – Foundations, Architectures, and Building AI Agents (8 Hours)
Session 1 (2 Hours): Introduction to Agentic AI (Module 1)
Learning Topics (1.5 hours):
- What is Agentic AI?
- Difference between Agentic AI and traditional AI models
- Applications of autonomous AI agents in business, healthcare, and finance
- Agentic AI vs. Generative AI: Key differences and overlaps
- Ethical, legal, and safety considerations in deploying autonomous AI
Hands-On (0.5 hours):
- Case Study: Analyze 3 existing Agentic AI frameworks (e.g., AutoGPT, Devin AI, BabyAGI).
- Group Discussion: Identify potential ethical risks and propose mitigation strategies for one real-world use case.
Session 2 (2 Hours): Fundamentals of AI Agents (Module 2)
Learning Topics (1.5 hours):
- Types of AI Agents: Reactive, Planning, and Learning agents
- Agent Architecture: Perception, Decision, and Action
- Belief-Desire-Intention (BDI) model overview
- State representation and memory (short-term vs. long-term)
Hands-On (0.5 hours):
- Implement a simple reactive agent in Python that follows environment-based rules.
- Design exercise: Create an architecture diagram for a goal-driven agent.
Session 3 (2 Hours): Building Autonomous AI Agents (Module 3)
Learning Topics (1.5 hours):
- Creating AI agents using LLMs (GPT, Claude, Gemini, etc.)
- Fine-tuning agents for specific tasks
- Prompt Engineering and Chain-of-Thought reasoning
- Managing agent context and task objectives
Hands-On (0.5 hours):
- Build a simple LLM-powered assistant that can summarize documents or respond contextually.
- Experiment with prompt templates and observe changes in agent reasoning.
Session 4 (2 Hours): Multi-Agent Systems (Module 4) + Reinforcement Learning (Module 5)
Learning Topics (1.5 hours):
- Understanding Multi-Agent Systems (MAS): Cooperative vs. Competitive setups
- Communication between AI agents (message passing, LLM-based coordination)
- Basics of Reinforcement Learning (RL) for autonomous decision-making
- Reward functions, policy learning, and Q-Learning overview
Hands-On (0.5 hours):
- Simulate a two-agent interaction (collaboration or negotiation).
- Implement a basic Q-Learning example (agent learns to navigate a simple grid environment).
Day 2 – Memory, Tool Use, Self-Learning, and Deployment (8 Hours)
Session 1 (2 Hours): Memory, Retrieval, and Planning (Module 6)
Learning Topics (1.5 hours):
- Adding memory to AI agents: Vector Databases and RAG (Retrieval-Augmented Generation)
- Hierarchical planning and task prioritization
- Tools and frameworks for autonomous task scheduling
Hands-On (0.5 hours):
- Integrate a memory layer using a vector database (e.g., FAISS or Chroma) into an LLM agent.
- Demonstrate persistent memory retrieval for multi-turn conversations.
Session 2 (2 Hours): Tool-Using AI Agents & API Integration (Module 7)
Learning Topics (1.5 hours):
- How AI agents use external APIs and databases
- LangChain, Auto-GPT, and related frameworks
- Automating workflows and business processes using Agentic AI
Hands-On (0.5 hours):
- Create an AI agent that calls an external API (e.g., weather or calendar API).
- Demonstrate how the agent uses tool outputs to complete a complex task autonomously.
Session 3 (2 Hours): Self-Learning & Real-World Applications (Modules 8 & 9)
Learning Topics (1.5 hours):
- How AI agents self-improve via meta-learning and self-evolution
- Feedback loops and automated fine-tuning strategies
- Industry applications: Business automation, healthcare assistants, research agents, trading bots
Hands-On (0.5 hours):
- Implement a simple feedback-based learning mechanism (agent improves based on user scores).
- Case Study: Analyze a self-learning agent used in finance or healthcare automation.
Session 4 (2 Hours): Deployment & Scaling (Module 10 + Final Project)
Learning Topics (1.5 hours):
- Deploying AI agents in cloud and edge environments
- Scaling for performance and reliability
- Monitoring and continuous optimization of deployed agents
- Governance and security in large-scale deployments
Hands-On (0.5 hours):
- Final Project:
○ Build and deploy a fully functional AI agent that performs an autonomous workflow (e.g., scheduling assistant, knowledge retrieval bot, or data summarizer).
○ Present design, architecture, deployment flow, and performance metrics.
Speaker Profile : Neerai is AI Consultant, Solution Architect & Trainer with 15+ years of experience in Generative AI, LLMs, AI Agents, Cloud-Native AI, and DevOps, specializing in Agentic AI Architectures and Autonomous Multi-Agent Systems. Expert in AI Agent Frameworks: LangChain, LangGraph, CrewAI, AutoGen – designing LLM-powered autonomous agents for decision-making, workflow automation, and self-reflective AI models.
Proven Expertise in RAG (Retrieval-Augmented Generation): Architected Corrective RAG, Self-Reflective RAG, and Self-Route RAG systems, leveraging LangGraph and cloud-based vector databases (ChromaDB, Pinecone, Weaviate, FAISS).Cloud & AI Model Integration Specialist: Deploying AI-driven automation on Azure OpenAI, AWS Bedrock, Google Vertex AI, integrating GPT-4o, Claude, Mistral, and open-source LLMs with LangChain-powered applications.
- Certified AI Cloud Trainer (AZ-305, AZ-400, AWS SA Pro, GCP Architect, CKA, AI/ML) with experience delivering 500+ training sessions on AI-Dev Integration, Multi-Agent Systems, and Autonomous AI Workflows.
- Google Certified Instructor (GCI), and SME for Data Engineering and Professional Architect.
- Databricks Certified Instructor (DCI) for all levels of Databricks certification and data engineering consultation projects.
- Lead SME for Generative AI skill development at Hexaware, TCS, and Wipro, with expertise in Python for GenAI, Prompt Engineering, and LLMs.
- Delivered solutions in Onsite & Offshore models across USA, Canada, Germany, South Africa, Australia,Singapore, UAE and India.
Who Should Attend:
- AI & ML Engineers
- Data Scientists & AI Enthusiasts
- Developers & Software Engineers
- Entrepreneurs & Product Managers
- Business Leaders Exploring AI Automation
