I am a student in a computer-related major at Dalian University of Technology.
I am more interested in how a feature can be implemented in practice. I usually develop my skills through complete projects rather than stopping at isolated algorithms or demos.
A multi-agent platform designed for software engineering learning and research scenarios.
The project integrates five types of Agents—paper reading, problem decomposition, result interpretation, code reproduction, and project planning—into a unified system, forming a basic closed loop across the frontend, backend, database, and Agent Runtime.
The current engineering components include:
- React + TypeScript frontend
- FastAPI unified API Gateway
- Supabase Auth, PostgreSQL, and RLS
- Conversation / Message data persistence
- Category-based Agent routing
- Signed WebSocket Agent communication
React · TypeScript · FastAPI · Supabase · PostgreSQL · WebSocket
A handwritten digit recognition project built with PyTorch, including data processing, CNN model construction, training and evaluation, and prediction result visualization.
Python · PyTorch · CNN
A path planning project based on A* Search, including map loading, obstacle processing, heuristic search, path reconstruction, and result visualization.
Python · A* · Path Planning
A regression experiment based on the California Housing dataset, including data preprocessing, model training, evaluation metric comparison, and experimental result analysis.
Python · scikit-learn · Regression
Interested in Agent task decomposition, routing, context management, and how multiple intelligent capabilities can be integrated into a unified application.
Interested in image classification, CNNs, and visual tasks, with the goal of continuing toward more complete vision applications.
Interested in API design, authentication, database architecture, service layering, and frontend-backend collaboration.
Interested in AI-assisted tools for paper understanding, code reproduction, experimental analysis, and other research scenarios.