Course overview
Course overview
Technical skills gained
Technical skills gained
- Build a solid understanding of JSON handling and HTTP/HTTPS communication in C++ using industry-standard libraries
- Construct AI API requests, parse structured responses, and handle errors correctly
- Learn the Builder Pattern, dependency injection, and composition through real projects
- Build a complete Retrieval-Augmented Generation pipeline from scratch in C++
- Design applications that support multiple AI backends with minimal code changes
Mindset and foundation
Mindset and foundation
- Understand what AI wrapper libraries are doing underneath
- Build a foundation that outlasts changing APIs, models, and tools
- Develop architectural thinking that applies well beyond this course
Key topics Covered
Key topics Covered
- OpenAI API integration from C++ using direct HTTP requests
- Ollama local AI model integration for free, private, offline AI
- JSON handling with nlohmann-json and HTTP/HTTPS with cpp-httplib
- Building a complete RAG system in C++
- The Builder Pattern with method chaining
- Progressive OOP: monolithic code to functions to full classes
- Dependency injection and composition in practice
- Structured output: prompting for JSON and parsing into C++ structs
- Cross-platform setup for Windows, macOS, and Ubuntu
- Embeddings, vector representations, and similarity search
WHAT MAKES THIS COURSE DIFFERENT
WHAT MAKES THIS COURSE DIFFERENT
- No wrapper libraries or pre-built AI SDKs. Every HTTP request, JSON payload, and parsed response is visible and explained. Students understand what is happening at every layer, not just how to call a function.
- Dual backend support throughout. All five projects support both OpenAI and Ollama with a simple configuration change, giving students full flexibility with no API cost required.
- Progressive OOP through a real project. The same chatbot is built three ways, monolithic, function-based, then fully object-oriented, so students see exactly why and when classes are an improvement.
- A complete RAG system in C++. RAG courses almost universally use Python. Building a full pipeline in C++ is rare and demonstrates C++ as a serious language for modern AI development.
- Five progressive projects with real architectural growth. Every project builds on the last, reusing and extending real code the way professional development actually works.