AI Engineering

AI Engineering: LLMs, RAG, AI Agents & Generative AI Tutorials | GeekyCodes

AI Engineering Tutorials

Build Production-Ready AI Systems

AI Engineering is the discipline of building reliable, scalable, and production-ready applications powered by artificial intelligence and large language models (LLMs).

Unlike traditional machine learning, modern AI applications often require much more than training a model.

A production AI system may involve:

LLMs → Prompt Engineering → Embeddings → Retrieval → Reranking → RAG → Agents → Evaluation → Observability → Production

At GeekyCodes, this AI Engineering tutorial series focuses on understanding these components from first principles and learning how they fit together to build real-world AI applications.

[Start with RAG Fundamentals →]

What Is AI Engineering?

AI Engineering combines concepts from:

  • Machine Learning
  • Deep Learning
  • Natural Language Processing
  • Large Language Models
  • Information Retrieval
  • Vector Databases
  • Software Engineering
  • Distributed Systems
  • MLOps

The objective is not simply to call an LLM API.

A production AI engineer needs to understand questions such as:

  • How do LLMs generate responses?
  • Why do LLMs hallucinate?
  • How can external knowledge be provided to an LLM?
  • How do embeddings represent meaning?
  • How does vector search work?
  • When should BM25 be used?
  • Why combine sparse and dense retrieval?
  • How does reranking improve retrieval quality?
  • When should RAG be used?
  • When should an AI agent be used?
  • How do we evaluate an LLM application?
  • How do we monitor and improve AI systems in production?

This learning path covers these concepts progressively.

AI Engineering Learning Path

AI Engineering
│
├── 1. LLM Fundamentals
│
├── 2. Prompt Engineering
│
├── 3. Embeddings
│
├── 4. Information Retrieval
│
├── 5. Vector Search
│
├── 6. RAG
│
├── 7. Reranking
│
├── 8. Advanced RAG
│
├── 9. AI Agents
│
├── 10. Agentic AI
│
├── 11. LLM Evaluation
│
├── 12. Production AI
│
└── 13. AI System Design

1. Large Language Models

Before building RAG systems and AI agents, it is important to understand how modern language models work.

Learn the fundamentals behind LLM-powered applications, including:

  • What are Large Language Models?
  • How LLMs Work
  • Tokenization
  • Embeddings
  • Attention Mechanism
  • Transformers
  • Positional Encoding
  • Context Windows
  • Inference
  • Temperature and Sampling
  • Prompt Engineering

Recommended Learning

[Explore LLM Fundamentals →]

2. Retrieval-Augmented Generation (RAG)

RAG is one of the most important architectures for building applications that combine LLMs with external knowledge.

Instead of relying entirely on the knowledge stored in model parameters, RAG retrieves relevant information from an external knowledge source and provides that information to the LLM as context.

RAG Fundamentals

Start here if you’re new to RAG.

Tutorials

  1. What is RAG?
  2. Why LLMs Hallucinate
  3. How RAG Works
  4. Dense vs Sparse Retrieval
  5. BM25 Explained
  6. Hybrid Search
  7. Cross-Encoder Reranking
  8. Parent-Child Retrieval
  9. Multi-Query Retrieval
  10. Self-RAG
  11. Corrective RAG
  12. Graph RAG
  13. Agentic RAG

Explore RAG Fundamentals

3. Embeddings & Vector Search

Embeddings are fundamental to modern AI applications.

They allow text, documents, images, and other data to be represented as vectors that capture semantic relationships.

Learn:

  • What Are Embeddings?
  • How Text Embeddings Work
  • Semantic Similarity
  • Cosine Similarity
  • Vector Search
  • Approximate Nearest Neighbor Search
  • Vector Indexes
  • Vector Databases
  • Metadata Filtering
  • Hybrid Vector Search

Vector Database Concepts

Explore concepts behind systems such as:

  • FAISS
  • Pinecone
  • Milvus
  • Chroma
  • Weaviate

[Explore Embeddings & Vector Search →]

4. Information Retrieval

A good RAG system depends heavily on retrieval quality.

Understanding information retrieval helps you design better search pipelines rather than treating retrieval as a black box.

Learn:

  • Information Retrieval Fundamentals
  • Sparse Retrieval
  • Dense Retrieval
  • BM25
  • TF-IDF
  • Semantic Search
  • Hybrid Search
  • Query Expansion
  • Query Rewriting
  • Metadata Filtering
  • Retrieval Evaluation
  • Reranking

[Explore Information Retrieval →]

5. Reranking

Retrieval and reranking solve different problems.

A retriever identifies potentially relevant documents, while a reranker can evaluate the retrieved candidates more carefully and reorder them.

Topics include:

  • What is Reranking?
  • Cross-Encoder vs Bi-Encoder
  • Why Reranking Improves RAG
  • Retrieve → Rerank → Generate
  • Reranker Selection
  • Reranking Strategies
  • Evaluating Retrieval Quality

[Explore Reranking →]

6. Advanced RAG

Basic RAG is only the beginning.

Real-world applications often require more sophisticated retrieval and reasoning strategies.

Explore:

  • Parent-Child Retrieval
  • Multi-Query Retrieval
  • Query Rewriting
  • HyDE
  • Self-RAG
  • Corrective RAG
  • Adaptive RAG
  • Graph RAG
  • Agentic RAG
  • Multi-Document RAG
  • Conversational RAG

The goal is to understand when each architecture should be used and what problem it solves.

[Explore Advanced RAG →]

7. AI Agents

AI agents extend LLM applications beyond simple question answering.

An agent can:

  1. Understand a goal
  2. Plan a task
  3. Select tools
  4. Execute actions
  5. Observe results
  6. Reason about the outcome
  7. Continue or modify the plan

Learn the fundamentals of:

  • What Are AI Agents?
  • Agent Architecture
  • Tool Calling
  • Function Calling
  • Agent Memory
  • Planning
  • Reflection
  • Tool Selection
  • Multi-Agent Systems
  • Agentic Workflows

[Explore AI Agents →]

8. Agentic AI & LangGraph

As AI applications become more complex, developers need frameworks for controlling multi-step workflows.

Explore concepts such as:

  • LangGraph Fundamentals
  • State Management
  • Nodes and Edges
  • Conditional Routing
  • Tool Calling
  • Human-in-the-Loop
  • Memory
  • Multi-Agent Architectures
  • Reflection
  • Planning
  • Agentic RAG
  • Production Agent Workflows

[Explore LangGraph & Agentic AI →]

9. LLM Evaluation

Building an AI application is only half the problem.

The difficult question is:

How do you know whether your AI application is actually working?

LLM applications require evaluation across multiple dimensions.

Learn about:

  • Response Quality
  • Faithfulness
  • Relevance
  • Context Precision
  • Context Recall
  • Hallucination Detection
  • Retrieval Evaluation
  • LLM-as-a-Judge
  • Human Evaluation
  • Evaluation Datasets
  • Regression Testing
  • Production Monitoring

[Explore LLM Evaluation →]

10. Fine-Tuning

RAG isn’t always the answer.

Sometimes a model needs to be adapted to a particular task, domain, format, or behavior.

Learn:

  • What is Fine-Tuning?
  • Pretraining vs Fine-Tuning
  • Instruction Tuning
  • Supervised Fine-Tuning
  • LoRA
  • QLoRA
  • PEFT
  • Dataset Preparation
  • Fine-Tuning Evaluation
  • Fine-Tuning vs RAG
  • When Should You Fine-Tune an LLM?

[Explore LLM Fine-Tuning →]

11. Production AI Engineering

A prototype that works locally is very different from a production AI system.

Production AI engineering requires consideration of:

  • Latency
  • Cost
  • Scalability
  • Reliability
  • Security
  • Observability
  • Evaluation
  • Prompt Management
  • Model Selection
  • Caching
  • Rate Limits
  • Failure Handling
  • Data Privacy
  • Governance

A typical production architecture might look like:

                  User
                   │
                   ▼
              Application
                   │
                   ▼
             AI Orchestrator
                   │
          ┌────────┴────────┐
          ▼                 ▼
      Retrieval          Tools
          │                 │
          ▼                 ▼
    Vector Database      APIs
          │
          ▼
       Reranker
          │
          ▼
          LLM
          │
          ▼
       Evaluation
          │
          ▼
       Response

[Explore Production AI Engineering →]

12. AI System Design

Learn how to design complete AI systems rather than individual components.

Topics include:

  • RAG System Design
  • AI Chatbot Architecture
  • Enterprise Search
  • Recommendation Systems
  • AI Agents
  • Multi-Agent Systems
  • LLM Serving
  • Vector Database Architecture
  • LLM Evaluation Systems
  • AI Observability
  • Scalable AI Infrastructure

[Explore AI System Design →]

🛠️ Hands-On AI Engineering Projects

The best way to learn AI Engineering is to build complete systems.

Project Ideas

Production RAG Chatbot

Build a RAG application with:

Documents
   ↓
Document Processing
   ↓
Chunking
   ↓
Embeddings
   ↓
Vector Database
   ↓
Hybrid Retrieval
   ↓
Reranking
   ↓
LLM
   ↓
Evaluation

[Build a RAG Chatbot →]

Agentic RAG System

Build an AI system that can decide when to search, retrieve information, use tools, and reason over results.

[Build Agentic RAG →]

Multi-Agent AI System

Build specialized agents for:

  • Planning
  • Retrieval
  • Reasoning
  • Tool Use
  • Reflection
  • Governance

[Build a Multi-Agent System →]

🎯 AI Engineering Interview Preparation

Preparing for an AI Engineer, GenAI Engineer, Machine Learning Engineer, or Data Scientist interview?

Explore questions covering:

  • LLMs
  • RAG
  • Embeddings
  • Vector Databases
  • Prompt Engineering
  • AI Agents
  • LangGraph
  • Fine-Tuning
  • LLM Evaluation
  • System Design
  • Production AI

[Explore AI Engineering Interview Questions →]

📚 AI Engineering Roadmap

If you’re starting from scratch, follow this order:

Python
  ↓
Machine Learning Fundamentals
  ↓
Deep Learning
  ↓
NLP
  ↓
Transformers
  ↓
Large Language Models
  ↓
Embeddings
  ↓
Information Retrieval
  ↓
Vector Search
  ↓
RAG Fundamentals
  ↓
Advanced RAG
  ↓
AI Agents
  ↓
Agentic AI
  ↓
LLM Evaluation
  ↓
Production AI
  ↓
AI System Design

[Follow the Complete AI Engineering Roadmap →]

🔗 Explore More

AI Engineering

[LLM Fundamentals]
[RAG Fundamentals]
[Embeddings & Vector Search]
[Information Retrieval]
[AI Agents]
[LangGraph]
[LLM Evaluation]
[Fine-Tuning]
[Production AI]
[AI System Design]

Related Learning Paths

[Data Engineering →]

[Machine Learning →]

[Interview Preparation →]

[Projects →]

Start Your AI Engineering Journey

Whether you’re learning Generative AI for the first time or building production-grade AI applications, the GeekyCodes AI Engineering learning path takes you from LLM fundamentals to RAG, AI agents, evaluation, and production AI systems.

Learn the concepts. Build the system. Understand how it works in production.

Start with RAG Fundamentals