Developing Advanced Generative AI Applications
Duration: 3 Days / 24 hrs
Introduction
This 3-day, hands-on course focuses on building advanced Generative AI applications that go beyond basic prompting. Participants learn how to fine-tune models efficiently, design memory-aware chatbots, work with multi-modal inputs, and build agentic, autonomous AI systems using modern frameworks.

Objectives
By the end of this course, participants will be able to:
- Apply parameter-efficient fine-tuning techniques like LoRA and IA3
- Build GenAI applications with memory, persistence, and long-context handling
- Implement tool calling, prompt chaining, and advanced output strategies
- Design agentic and autonomous AI systems using industry-grade frameworks

Key Takeaways
Participants will leave with:
- Hands-on experience building production-grade GenAI systems
- Practical understanding of fine-tuning, multi-modal prompting, and orchestration
- Confidence to design scalable, secure, and compliant AI applications

Training Methodology (Learning by Doing)
- Hands-on labs across all modules—no slide-only sessions
- Incremental builds from fine-tuning to agentic workflows
- Real-world scenarios including risk, cost, and compliance considerations
- Build, test, refine with continuous evaluation and debugging

Course Outline
Parameter-Efficient Finetuning Techniques
- Comparison with traditional finetuning
- Prompt-based methods
- Low-Rank Adaptation (LoRA), IA3, and others
- Performance evaluation
Chatbots with Memory and Persistence
- Persisting messages
- Handling long conversations
- Cross-session memory
- Fact extraction and privacy considerations
Multi-modal Prompting and Interaction
- Processing non-textual inputs
- Prompting with images and text
- Stable Diffusion overview
- Cost optimization
- Use cases
Streaming, Ensembling, and Output Techniques
- Streaming LLM outputs
- Self-consistency and post-processing
- Mixture methods and self-refinement
Function and Tool Calling
- Binding functions to LLMs
- Handling responses and errors
- Security and retries
- LangChain Toolkits
Advanced Use Cases
- Embedded agents
- Agent-based modeling
- Code generation and cybersecurity
- Risk management and EU AI Act
Prompt Chaining and Directed Graphs
- Predetermined and generated paths
- Context passing
- Conditional routing and debugging
Agentic AI and Autonomous Systems
- Definitions and characteristics
- Comparison with traditional AI
- Real-world applications
- Architectures: Reactive, Deliberative, Hybrid
- Frameworks: AutoGen, LangGraph, CrewAI