Jiang Bin
About Me
AI Application Engineer / Python Backend Developer
I am an AI Application Engineer / Python Backend Developer based in Hangzhou / Wuhan, with 15 years of experience in software development and testing, spanning Internet, Telecom, and Industrial sectors. From embedded hardware and automation testing to cloud backend and LLM applications, I excel at transforming complex business requirements into stable, scalable engineering solutions.
In recent years, I have focused on: RAG (Retrieval-Augmented Generation) systems, Defect Detection LLM applications, Python backend development, and Azure cloud deployment. I specialize in planning and implementing technical solutions from 0 to 1, including API design, database modeling, performance optimization, and engineering operations.
Skills
AI / LLM & RAG
- RAG: Document parsing, Vector retrieval (Milvus), Agent RAG
- Prompt Engineering: Multi-turn dialogue, Tool calling, Text-to-SQL
- Defect Detection LLM: YOLOv8 + General LLM collaboration
- LLM Function Calling: Workflow orchestration, Vanna.ai SQL generation
- Deep Learning: PyTorch 2.2.0+, TensorFlow 2.19.1, Keras 3.11.2
Backend & Data
- Languages: Python (FastAPI, Django, Flask), C++
- Async: Celery, RabbitMQ, WebSocket, Async task scheduling
- DB: MariaDB, MySQL, Milvus, PostgreSQL, SQLAlchemy 2.0+
- Data: Scrapy, Pandas, NumPy, SciPy, UMAP, HDBSCAN clustering
- API Design: RESTful API, Pydantic 2.9+, Async HTTP
Cloud Native & Engineering
- Azure: Function, ML Service, Serverless, Web App, Blob Storage
- DevOps: Jenkins, GitLab CI, GitHub Actions, Docker, JIRA
- ML DevOps: Wheel package auto-build, GitHub Actions CI/CD, Model distribution
- Test: RobotFramework, pytest, unittest
- Embedded: C / Assembly (MCU, ARM, DSP)
Projects
Defect Detection LLM Platform
- Backend Arch: Built backend for defect detection system, integrating visual models (YOLOv8) & LLMs for online identification.
- API Design: Designed RESTful APIs using FastAPI & Pydantic 2.9+ models for high efficiency and stability.
- Engineering: Implemented Function Calling & ML model deployment (Local/Cloud), optimizing resource allocation, supporting Azure ML Endpoint.
- Performance: Designed RabbitMQ async queues & WebSocket communication to boost throughput and UX.
- Tech Stack: PyTorch 2.2.0+, ONNX Runtime 1.16.0+, Celery, RabbitMQ, Azure Web App, Blob Storage
RAG Intelligent Q&A System
- Full Stack: Responsible for backend architecture, coding, workflow, and deployment (Pre-launched).
- Core Tech: Led AI knowledge base construction, Prompt Engineering, and RAG development.
- Data Opt: Designed Milvus vector database structure, significantly improving retrieval efficiency and accuracy.
- Text-to-SQL: Implemented natural language to SQL conversion, integrated Vanna.ai for intelligent queries.
- Tech Stack: FastAPI, Milvus, PostgreSQL, SQLAlchemy 2.0+, Agent RAG
Quality Analysis Platform (piyi-api)
- Algorithm Integration: Developed quality analysis models based on TensorFlow 2.19.1 + Keras 3.11.2, implementing UMAP dimensionality reduction + HDBSCAN clustering.
- Feature Engineering: Used NumPy, SciPy, Pandas for feature extraction and data processing, supporting high-dimensional data visualization.
- Hybrid Deployment: Supported automatic switching between local models and Azure ML Endpoint for elastic scaling.
- Tech Stack: TensorFlow 2.19.1, Keras 3.11.2, UMAP-learn 0.5.9, HDBSCAN 0.8.40, scikit-learn 1.7.2
AI Inference Engine (AI-Project)
- Modular Design: Developed independently installable Python packages (ai_libs + fabric_defect), supporting one-click pip deployment.
- High Performance: Implemented YOLOv8 object detection based on PyTorch 2.2.0+ and ONNX Runtime 1.16.0+, supporting GPU acceleration.
- CI/CD Automation: Configured GitHub Actions for automatic wheel package building and GitHub Release, achieving production-grade deployment.
- Tech Stack: PyTorch 2.2.0+, ONNX 1.15.0+, OpenCV 4.8.1+, pyproject.toml, GitHub Actions
CASPI Business System (caspi-api)
- API Development: Developed core business APIs based on FastAPI, implementing file upload, batch deletion, and other functions.
- CI/CD Optimization: Refactored GitHub Actions workflows, improving automated deployment efficiency.
- Continuous Iteration: 22 PRs, high activity, excellent code quality (83% merge rate).
- Tech Stack: FastAPI, Pydantic, GitHub Actions, Azure
Genesis Large-Scale Project
- Agent RAG Practice: Combined AI agents with retrieval-augmented generation to handle complex business logic.
- Large-Scale System: Project volume 174MB, multi-module, multi-business line comprehensive system development.
- Continuous Iteration: 25 high-quality code commits, rich system architecture design experience.
- Tech Stack: Python, Agent RAG, Complex Business Systems
GitHub Open Source Contributions
Contribution Stats
156+
Code Commits
11
Active Repos
30+
Pull Requests
Top Projects
1 defect-detection-trigger-task 29 commits
2 genesis 25 commits
3 caspi-api 20 commits
4 defect-detection-api 17 commits
5 genesisUserManagement 13 commits
📊 Data source: GitHub Contributions Summary
Experience
Hangzhou Clayton Industries · AI Application Engineer
- Responsible for backend design & dev using Pydantic + FastAPI, ensuring system stability.
- Designed intelligent Q&A system, focusing on Prompt Engineering & RAG technology selection.
- Developed defect detection ML algorithms and engineered LLM applications.
- Responsible for Azure cloud system engineering (Backend, Function Calling, ML) and performance optimization.
Hangzhou C&U Group · Python Backend Developer
- Responsible for technical planning of the MES platform backend module.
- Implemented business logic (Django), designed async Celery tasks & MariaDB structure.
- Participated in API design & review, wrote technical documentation & interface standards.
Nokia Solutions and Networks · Test Development Engineer
- Automation Test: Projects based on RobotFramework (Grey-box) and TTCN-3.
- Software Dev: Participated in C++ software development and testing.
- Tool Dev: Developed automation & platform tools using Python, Shell, Lua.
- DevOps: Practiced CI/CD using Jenkins, GitLab CI, Docker, etc.
Hangzhou Geling Technology · Python Developer
- Built data cloud platform based on Python Django.
- Responsible for web crawler data collection and processing.
Xizi YOUMY · Software Development Engineer
- Responsible for logic software programming based on PLC, MCU, ARM, DSP (C/ASM).
- Developed and maintained test fixtures for embedded systems; conducted safety system certification.
- Provided technical guidance for pioneer products and resolved on-site issues.
- Achievements: Completed SIL 2 certification; released the first batch of new national standard escalators.
Education
- South-Central Minzu University · Bachelor · Biomedical Engineering · 2005 – 2010
- CET-6
Volunteer & Community
West Lake Maker Space · Mentor
Served a community of 1000+ young makers, sharing technical experience and guiding project development; organized and hosted multiple technical workshops and hackathons.
1566 Maker Charity Fund · Ambassador
Provided charity funds and technical support for potential maker projects; participated in project reviews to help select and cultivate innovative projects with social value.
Contact Me
Welcome to connect for collaboration on Python / AI Engineering / RAG Systems / Industrial Intelligence.
- Email: jiangbingo@hotmail.com
- GitHub: github.com/jiangbingo
- Location: Hangzhou / Wuhan, China
