Software Engineering student at Jiangxi Agricultural University.
Current direction: Python Backend -> AI Agent Backend
Current focus: FastAPI, testing, clean architecture, RAG/Agent engineering
I use GitHub as a public evidence board: runnable projects, clear README files, tests, and project retrospectives.
fastapi-notes-crud
A small but production-structured FastAPI Notes CRUD API.
What it shows:
- layered backend design: route -> dependency wiring -> service -> repository -> models
- Pydantic v2 request/response validation
- search and pagination
- TestClient / pytest checks
- README with API examples and local run steps
python-backend-ai-learning
Public learning trail for Python Backend -> AI Backend.
What it shows:
- daily learning reviews
- backend notes
- project retrospectives
- roadmap from FastAPI to SQLAlchemy, Redis, pytest, RAG, and Agent backend
| Time | Focus | Output |
|---|---|---|
| May | FastAPI CRUD, layered architecture, TestClient | fastapi-notes-crud |
| June | SQLAlchemy, database design, Redis, pytest | database-backed API project |
| July | RAG and Agent basics | AI knowledge-base backend |
| August | Complete AI backend project | portfolio-ready AI Agent app |
- Languages: Python, Java, C/C++, JavaScript
- Backend: FastAPI, Spring Boot, REST API
- Data: Pydantic, SQLAlchemy, Redis
- Testing: pytest, FastAPI TestClient
- Tools: Git, Linux, VS Code, Claude Code, Codex
- Learning Track: LeetCode Hot100, CS fundamentals, postgraduate CS preparation
My AI-assisted development workflow:
task brief -> architecture boundary -> small implementation -> validation -> review -> GitHub evidence
I try to keep AI as the executor, while I keep ownership of requirements, architecture boundaries, review, and explanation.
I want to become a backend engineer who can use AI tools without losing engineering judgment:
- define real problems
- design maintainable systems
- verify results with tests
- write clear documentation
- explain tradeoffs in interviews
Private repositories contain environment backups, raw learning vaults, and personal AI workflow configuration. Public repositories are kept as interview-facing evidence.