This repository contains my learning notes, code examples, and experiments while studying modern AI systems engineering.
The goal of this repository is not to build a full framework.
Instead, I want to better understand how modern AI applications actually work behind the scenes.
Topics include:
- async workflows
- queues and streaming
- tool calling
- structured outputs
- multi-agent systems
- verification and evaluation
- benchmarking
- provider abstraction
- workflow orchestration
Most examples are intentionally small and simplified.
I prefer learning concepts through minimal working examples and real workflow understanding instead of only memorizing framework APIs.
ai-systems-notes/
│
├── README.md
│
├── 01_async/
├── More is coming..
Each folder usually contains:
- minimal code examples
- notes.md
- small experiments
- workflow sketches
01 Async check details
02 Queue Streaming check details
03 Structure Output check details
More is coming..
While learning agent systems, I realized that modern AI applications are very different from traditional ML projects.
Many systems involve:
- async operations
- orchestration layers
- streaming
- tool execution
- verification pipelines
- long-running workflows
At first, many of these concepts felt confusing.
This repository is my attempt to gradually build a clearer understanding through notes and experiments.
These notes are part of my personal learning process.
Some examples are intentionally simplified in order to focus on concepts and workflow understanding.