Generative AI (GenAI)

• • Foundations of LLMs & Prompt Engineering
    • Fundamentals of LLMs
      • Understand how LLMs work, including tokenization and embeddings.
      • Explore how models like GPT interpret and generate text.
      • Project: Visualize and compare embeddings.
    • Prompt Engineering in Practice
      • Learn how to structure and format prompts effectively.
      • Understand chat roles: system, user, assistant.
      • Practice designing multi-turn prompts and chaining logic.
      • Project: Create a multi-turn chatbot.
    • Prompt Engineering with OpenAI API
      • Apply prompting techniques for summarizing, classifying, and transforming text.
      • Use OpenAI API to test and iterate prompts programmatically.
      • Project: Implement and compare zero-shot, one-shot, and few-shot prompts for text summarization, sentiment detection, and email drafting.
    • Building Systems using Generative AI
      • Apply prompt engineering to solve real-world text tasks.
      • Split complex tasks into a pipeline of subtasks using multi-stage prompts.
      • Evaluate LLM inputs and outputs for safety, accuracy, and relevance.
      • Project: Build a customer-support agent that can flag, retrieve, and answer user queries.

• • LangChain & Agent Systems
    • Application Development using LangChain
      • Understand the core LangChain framework.
      • Learn how to use your own data with RAG (Retrieval-Augmented Generation).
      • Build chatbots powered by structured chaining logic.
      • Project: Create a document Q&A assistant using LangChain and your own PDFs.
    • Multi-Agent Systems with AutoGen
      • Learn to scale LLM performance through Agentic AI systems using AutoGen.
      • Compare agent patterns such as coordinator-worker, role-based agents, and tool-invoking agents.
      • Explore real-world task automation like document processing or customer service escalation.
      • Optional: Contrast AutoGen with CrewAI for understanding trade-offs.
      • Project: Build a 3-agent AutoGen system to automate customer support (intake, triage, escalation).
    • Agentic Design with LangGraph
      • Learn to design structured agent workflows with LangGraph.
      • Combine LangGraph’s visual DAG-based control with AutoGen’s conversational memory.
      • Project: Implement a LangGraph/AutoGen hybrid agent that dynamically routes tasks and retries failed steps.
    • Reasoning with LLMs
      • Learn how reasoning models work and when to use them.
      • Use chain-of-thought and self-reflection prompting techniques.
      • Apply reasoning to code and vision-based tasks.
      • Project: Use reasoning prompts to debug code and answer image-based questions step-by-step.

• • Reasoning Models & Advanced Agents
    • Advanced Use Cases with AutoGen
      • Design collaborative multi-agent systems using AutoGen’s planner-executor framework.
      • Integrate diverse LLMs and tools, support memory/context sharing, and enable agents to reason, critique, and self-correct.
      • Inject domain-specific logic, external APIs, and feedback loops for robust automation.
      • Project: Build an AutoGen-based resume screening system that ranks candidates using dynamic criteria.
    • Build with Anthropic – Computer Use
      • Understand Anthropic family of models, best practices, and multi-modal use cases.
      • Implement low-latency and cost-saving techniques like prompt caching.
      • Learn multi-modal and tool-augmented use cases.
      • Project: Build an AI assistant that opens a file, summarizes content, and suggests edits.
    • Reasoning + Agent Project Sprint
      • Apply reasoning and agent collaboration in a high-impact use case.
      • Combine LangGraph for workflow structuring and AutoGen for autonomous reasoning agents.
      • Project: Build a “Virtual Analyst” that reasons over documents, answers queries, and summarizes trends.

• Memory, Context Protocols & AI Safety
  • Long-Term Memory with LangGraph + AutoGen
    • Build agents with evolving memory using AutoGen’s user memory modules and LangGraph’s persistent nodes.
    • Support personalization and context continuity across sessions.
    • Project: Extend AutoGen agents with long-term adaptive memory.
  • Model Context Protocol (MCP)
    • Learn how MCP standardizes tool and data access across apps.
    • Connect multiple AI agents and services using a shared protocol.
    • Deploy and manage MCP servers and clients.
    • Project: Deploy an MCP server and connect mini-apps that share user context.
  • Guardrails for AI Systems
    • Recognize common failure modes in LLM systems.
    • Use guardrails to validate and filter model inputs and outputs.
    • Integrate I/O controls into RAG and agent-based apps.
    • Project: Add guardrails to a Generative AI app to block unsafe, biased, or irrelevant responses.
  • AI Project Best Practices – Building and Deploying End-to-End Applications
    • Learn the lifecycle of AI product development: scoping, prototyping, testing, and deployment.
    • Explore trade-offs: prompting vs. fine-tuning, cost vs. accuracy, latency vs. interactivity.
    • Understand fine-tuning fundamentals and implications.
    • Learn best practices for versioning, observability, feedback collection, and human-in-the-loop integration.
    • Evaluate AI systems for quality, safety, and robustness.