Lộ trình học tập AI Engineer được phân chia theo 12 chuyên mục:
01. Introduction
- Introduction
- What is an AI Engineer?
- AI Engineer vs ML Engineer
- Impact on Product Development
- Roles and Responsiblities
02. LLM Fundamentals
- Large Language Model (LLM)
- Inference
- Training
- Vector DBs
- RAGs
- AI vs AGI
- Pre-trained Models
- Tokens
- Context
- How LLMs Work
- Closed vs Open Source Models
- Type of Models
- Self-Hosted Models
03. Prompt Engineering
- Prompt Engineering
- ReAct
- CoT
- Zero-Shot
- Top-K
- Top-P
- Few-Shot
- Context Engineering
- Input Format
- System Prompting
- Role & Behavior
- Constraints
- Structured Output
- Function Calling
- Prompt Caching
- Streaming Responses
- Repetition Penalties
- Sampling Parameters
- Temperature
- External Memory
- RAG & Dynamic Filters
- Context Compaction
- Context Isolation
04. RAG & Vector Databases
- What are Embeddings
- Semantic Search
- Recommendation Systems
- Anomaly Detection
- Data Classification
- Open AI Embeddings API
- Gemini Embedding
- Jina
- Sentence Transformers
- Models on Hugging Face
- Vector Databases
- Purpose and Functionality
- Chroma
- Pinecone
- Weaviate
- FAISS
- LanceDB
- Qdrant
- Supabase
- MongoDB Atlas
- Indexing Embeddings
- Performing Similarity Search
- What are RAGs?
- RAG Usecases
- RAG vs Fine-tuning
- Chunking
- Embedding
- Vector Database
- Retrieval Process
- Generation
- Using SDKs Directly
- Langchain
- Llama Index
- Embedding Models
- Embeddings
05. Fine-Tuning
06. AI Agents
- AI Agents
- Tools & Function Calling
- Agents Usecases
- ReAct Prompting
- Manual Implementation
- OpenAI AgentKit & Agent SDK
- Claude Agent SDK
- Multi-agents
07. Model Context Protocol (MCP)
- Model Context Protocol (MCP)
- MCP Host
- MCP Server
- MCP Client
- Data Layer
- Transport Layer
- Building an MCP Server
- Building an MCP Client
- Connect to Local Server
- Connect to Remote Server
08. Multimodal AI
- Multimodal AI
- Multimodal AI Usecases
- Image Understanding
- Image Generation
- Video Understanding
- Audio Processing
- Text-to-Speech
- Speech-to-Text
- OpenAI Vision API
- DALL-E API
- Whisper API
- Hugging Face Models
- LangChain for Multimodal Apps
- LlamaIndex for Multimodal Apps
09. Frameworks & Tools
- Hugging Face
- LM Studio
- Hugging Face Hub
- Hugging Face Inference SDK
- Transformers.js
- Ollama
- Development Tools
- Haystack
- RAGFlow
- Hugging Face Tasks
10. Models & APIs
- Anthropic Claude
- Google Gemini
- OpenAI (GPT, o-series)
- Meta Llama
- Mistral
- Cohere
- OpenAI Response API
- Google Gemini APi
- DeepSeek
- Vertex AI Agent Builder
- Google ADK
- Choosing the Right Model
- Gemma
- Qwen
- OpenRouter
- Claude Messages API
- OpenAI-compatible APIs
- NanoBanana API
11. Evaluation, Safety & Ethics
- AI Safety and Ethics
- Prompt Injection Attacks
- Bias and Fairness
- Security and Privacy Concerns
- Conducting adversarial testing
- Content Moderation APIs
- Adding end-user IDs in prompts
- Robust prompt engineering
- Know your Customers / Usecases
- Constraining outputs and inputs
- LLM Observability
- LLM Evaluations
- Tracing & logging
- Cost/latency monitoring
- Production monitoring
- LangSmith
- Langfuse
- Helicone
- Arize AI
- Deterministic Evals
- Model-Based Evals
- Human Evals
- Evaluation Metrics
- Regression Testing
- DeepEval
- RAGAS