Course Outline
Day 1 – Generative AI & LLM Fundamentals
Overview of generative AI and LLM application scenarios
Understanding transformer-based architectures (GPT, LLaMA, T5, etc.)
Tokens, tokenization processes, and embeddings
Interacting with pre-trained models through APIs (OpenAI, Claude)
Accessing pre-trained models via Hugging Face
Prompting essentials: zero-shot and few-shot techniques
Practical exercise: prompt engineering in a Python environment
Developing a basic LLM-powered application (CLI or web interface)
Recognizing practical constraints: token limits, rate restrictions, and basic reliability strategies
Day 2 – RAG and Vector Search
The rationale behind RAG: augmenting LLMs with proprietary data
RAG architecture components: ingestion, indexing, retrieval, and generation
Document preparation and chunking strategies for retrieval
Generating text embeddings using APIs or Hugging Face
Introduction to vector databases (e.g., Chroma, Pinecone)
Practical exercise: constructing a semantic search script
Practical exercise: building a document Q&A system using RAG
Scaling ingestion and embedding processes (overview of large-scale workflows)
Design trade-offs in RAG: chunking methods, top-k selection, and balancing cost versus quality
Day 3 – Workflows, Agents, and Production
Understanding AI agents and their appropriate use cases
Introduction to LangGraph and graph-based LLM workflows
Practical exercise: creating a simple LangGraph workflow with integrated tools
Incorporating memory and multi-step reasoning into workflows
Integrating RAG with agents (agentic RAG)
Monitoring and evaluating LLM and RAG system performance
LLM application deployment options (APIs, containers, services)
Strategies for optimizing cost and performance
Fundamentals of safety, guardrails, and responsible AI usage
Capstone mini-project: end-to-end RAG/agent application demo
Requirements
Strong Python programming proficiency and familiarity with APIs.
Target Audience:
This course is designed for organizations aiming to transition from exploratory projects to functional LLM-powered solutions. It is ideal for software, backend, and full-stack engineers integrating LLMs into products; data and machine learning engineers focused on RAG, embeddings, and vector search; solution and enterprise architects planning LLM-based infrastructures; as well as technical product owners and engineering leaders tasked with assessing AI use cases, associated costs, and potential risks.
Testimonials (2)
The interactive style, the exercises
Tamas Tutuntzisz
Course - Introduction to Prompt Engineering
A great repository of resources for future use, instructor's style (full of good sense of humor, great level of detail)
