Generative AI

This module focuses on building and understanding the key architectures and models that power Generative AI systems.

• Unimodal Models: Architectures for single-modality tasks such as text-only or image-only applications.
• Multimodal Models: Techniques for integrating and processing data from multiple modalities.
• Transformers and Attention Mechanism: Core concepts, architecture, and practical applications.
• LLM Fine-Tuning: Adapting large language models for domain-specific use cases.
• LLM Evaluation: Metrics and methodologies to assess model performance.

Learn how to apply Generative AI in real-world settings using APIs and open-source frameworks.

• API-Based Usage: Platforms such as Hugging Face and OpenAI for seamless integration.
• Non-API-Based Usage: Open-source tools for building and customizing LLMs.
• Hands-On Exercises: Interactive labs for deploying text, image, and multimodal applications.

This module delves into the art of crafting effective prompts for Generative AI models: Prompt Optimization: Strategies for task-specific prompts.

• Code-Based and No-Code Approaches: Tools and techniques for advanced prompt engineering.
• Few-Shot and Chain-of-Thought Prompting: Advanced methodologies for improved performance.

This module explores advanced applications of Generative AI, including retrieval-augmented systems and autonomous agents.

• RAG Systems: Combining retrieval and generation for context-aware responses.
• AI Agents: Designing intelligent agents capable of autonomous actions.
• Platform Integration: Using frameworks like LangChain and vector databases.

This module focuses on developing robust, production-ready applications with Generative AI, and later on deploying and managing them in production environments.

• Planning Agents: Designing and implementing agents for complex workflows.
• User Interfaces for GenAI Applications: Best practices for building intuitive UIs.
• Case Studies: Real-world applications of agentic AI systems.
• Hosting AI Models: Best practices for deployment and scaling.
• LLMOps: Operational practices for managing large language models effectively.
• Security and Performance Monitoring: Ensuring robustness and reliability.

Participants will demonstrate their expertise through a comprehensive project

• Design and implement an end-to-end Generative AI solution.
• Incorporate advanced techniques such as RAG, multi-modal models, or agentic AI.
• Present the project with detailed documentation and a live demonstration.