Basic AI Fundamentals Syllabus

1. Header Information

Field	Value
Technical Group	AI Engineering
Topic Name	Basic AI Fundamentals
Topic Code	BAI_01
Version	1.0
Training Audience	Freshers / Interns with basic programming knowledge (Python)

2. Course Objectives

This topic introduces the fundamental concepts of Artificial Intelligence (AI) and Generative AI (GenAI), with a deep dive into Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). Trainees will learn the theoretical foundations of RAG, modern RAG architectures, and how to implement them using the LangChain framework. The course culminates in building a practical RAG agent to answer questions based on internal policy documents.

[!NOTE] This topic equips trainees with the necessary skills to understand and build modern AI applications using LLMs and RAG, adapting them to real-world enterprise scenarios like FPT's internal systems.

Output Standards table

Name	Code	Description
AI Concepts	OS1	Understand the core concepts of AI, GenAI, and the capabilities/limitations of LLMs.
RAG Theory	OS2	Comprehend the theoretical foundations and architecture of RAG (Indexing, Retrieval, Generation).
LangChain	OS3	Master the core components of the LangChain framework (Loaders, Splitters, Embeddings, Vector Stores).
Implementation	OS4	Apply knowledge to build a functional RAG Agent using LangChain and a Vector Database.

Detailed Learning Outcomes (LOs)

• AI & GenAI Fundamentals

  • OS1 LO-1 - Define Artificial Intelligence, Generative AI, and Large Language Models (LLMs).
  • OS1 LO-2 - Identify key capabilities of LLMs (Text Generation, Summarization, etc.) and their limitations (Hallucinations, Knowledge Cutoff).

• RAG Theoretical Foundations

  • OS2 LO-3 - Explain the concept of Retrieval-Augmented Generation (RAG) and why it is needed.
  • OS2 LO-4 - Differentiate between Parametric Memory (Model Weights) and Non-Parametric Memory (External Knowledge).
  • OS2 LO-5 - Compare Original RAG (Fine-tuning) vs. Modern RAG (In-Context Learning).

• Modern RAG Architecture

  • OS2 LO-6 - Describe the Indexing Phase: Document Loading, Text Splitting (Chunking strategies), and Embedding.
  • OS2 LO-7 - Explain Retrieval techniques: Vector Search, Hybrid Search (Dense + Sparse), and Re-ranking (Cross-Encoder).
  • OS2 LO-8 - Understand Generation strategies: Context Stuffing, Context Selection/Compression, and Prompt Engineering (Zero-shot, Few-shot, CoT).

• LangChain Framework

  • OS3 LO-9 - Utilize LangChain Document Loaders and Text Splitters to prepare data.
  • OS3 LO-10 - Implement Embeddings and Vector Stores (e.g., FAISS, Chroma) for data storage.
  • OS3 LO-11 - Configure Retrievers for semantic search.

• Practical Implementation

  • OS4 LO-12 - Build a complete RAG pipeline: Loading -> Splitting -> Embedding -> Storing -> Retrieving -> Generating.
  • OS4 LO-13 - Develop a ReAct Agent using LangChain to answer questions from specific document sets (e.g., FPT Policy).

3. Topic Outline

• Unit 01: Introduction to AI & Generative AI
• Unit 02: Introduction to RAG & Theoretical Foundations
• Unit 03: Modern RAG Architecture (Indexing, Retrieval, Generation)
• Unit 04: LangChain Framework & Core Components
• Unit 05: Building RAG Agent Using LangChain

4. Time Allocation

Activity Type	Percentage	Description
Concept/Lecture	40%	Concepts, theory (Intro, Architecture, LangChain)
Assignment/Lab	40%	Hands-on Labs (LangChain setup, RAG Agent build)
Guides/Review	10%	Code reviews, Q&A sessions
Test/Quiz/Exam	10%	Daily Quizzes, Final Project

5. Training Materials & Environments

Textbooks & Guides

• Internal Knowledge Base:
  • Introduction to AI & Generative AI
  • Introduction to RAG & Theoretical Foundations
  • Modern RAG Architecture
  • LangChain Framework and Core Components
  • Building RAG Agent Using LangChain

References

• LangChain Documentation
• RAG Paper (Lewis et al., 2020)
• Chain-of-Thought Prompting

Technical Requirements

• Python 3.9+
• Jupyter Notebook / VS Code
• OpenAI API Key (or equivalent LLM provider)
• Libraries: langchain, langchain-openai, faiss-cpu (or chromadb), pypdf

6. Assessment Scheme

Component	Quantity	Weight (%)	Note
Quiz	5	20%	Daily quizzes after each unit
Assignments	2	30%	Lab exercises (Data loading, Vector Store setup)
Final Project	1	50%	Build a RAG Agent for FPT Policy

Pass Criteria

• Total topic GPA >= 70/100
• Completed 100% of Quizzes and Assignments
• Successful demo of the Final Project (RAG Agent)

7. Training Delivery Principles

Role	Responsibility/Criteria
Trainees	Completion of Python basics. Active participation in labs.
Trainer	Senior AI Engineer / Tech Lead with RAG experience.
Training	Daily theory sessions followed by practical coding labs.
Re-Test	Allowed once for the Final Project if GPA < 70 but >= 50.
Marking	Code quality, functionality, and understanding of concepts.