Crafting Custom Agentic AI Solutions
Duration: 3 Days / 24 hrs
Introduction
This 3-day hands-on course focuses on designing and building custom agentic AI solutions that can reason, plan, use tools, and operate with context. Participants move beyond basic GenAI apps to create intelligent agents using modern frameworks like AutoGen, LangGraph, and CrewAI.
Objectives
By the end of this course, participants will be able to:
- Understand agentic AI concepts and architectures
- Design agents with memory, reasoning, and tool-use capabilities
- Implement planning and decision-making workflows
- Build, test, and optimize real-world single-agent AI applications
Key Takeaways
Participants will leave with:
- A strong mental model of agentic AI systems
- Hands-on experience building custom AI agents
- Practical patterns for memory, tool use, and reasoning
- Confidence to design production-ready agentic solutions
Training Methodology (Learning by Doing)
- Hands-on from Day 1 with guided coding and labs
- Incremental builds—agents evolve feature by feature
- Real-world scenarios for planning, reasoning, and recovery
- Debug, refine, repeat to mirror real development workflows
Course Outline
Introduction to Agentic AI
- Definition and characteristics
- Comparison with traditional AI and RPA
- Real-world applications and sector impact
Agent Architectures
- Reactive, deliberative, and hybrid models
- Frameworks: AutoGen, LangGraph, CrewAI
Memory & Context Management
- Types: Working, long-term, episodic
- Challenges and advanced techniques
Tool Use & Function Calling
- OODA loop: Observe, Orient, Decide, Act
- Direct vs indirect tool use
- Validating inputs/outputs
Planning & Reasoning
- Rule-based, goal-based, utility-based planning
- Deductive, inductive, abductive reasoning
- Handling uncertainty and errors
Building Single-Agent Applications
- Best practices for architecture and framework selection
- Testing, debugging, and optimization
- Integration patterns and recovery strategies
