Find the competitions that actually fit you.
面向大学生的竞赛发现、匹配与参赛记录平台
Live Product · 中文 · English
Stellula 是一个面向大学生的竞赛发现平台。
它想解决的不是“网上有没有竞赛信息”，而是一个更具体的问题：
我学这个专业，现在到底有哪些比赛适合我？
竞赛信息并不少，但它们常常散落在不同的官网、学校通知、公众号、社交平台和往届经验里。对于刚开始接触竞赛的学生来说，真正困难的往往不是找到一个比赛名称，而是判断：
- 这个比赛和我的专业有没有关系？
- 我现在的年级能不能参加？
- 为什么它会被推荐给我？
- 哪个链接才是官方入口？
- 我参加过什么，之后还能不能找回来？
Stellula 围绕这条学生路径设计：
发现 → 判断 → 参赛 → 记录 → 回看
它希望把“找比赛”从一次临时搜索，慢慢变成一份属于自己的大学竞赛轨迹。
根据专业、年级、兴趣与地区等信息，帮助学生缩小搜索范围，并优先展示更相关的竞赛。
推荐不应该只是一个黑盒结果。Stellula 尽量保留结构化的匹配依据，让学生能够理解一个比赛为什么出现在结果中。
赛事页面整理比赛简介、适合方向、参赛条件及官方资源。Stellula 不替代官方报名渠道，而是帮助用户更快抵达可信的信息来源。
用户可以记录自己想参加、已经参加以及完成过的比赛，让几年大学经历不必只剩下一张成绩单。
竞赛墙更接近一条个人经历时间线，而不是排行榜。
- 先有用，再聪明。 基础匹配逻辑应该独立成立，AI 不是前提。
- 官方来源优先。 重要赛事信息尽可能指向可验证的官方来源。
- 推荐需要解释。 用户应该知道“为什么是它”。
- 为普通学生设计。 不假设用户已经熟悉竞赛体系。
- 经历本身值得被记录。 奖项重要，但参赛经历不应该被简单压缩成“成功 / 失败”。
- 隐私默认收紧。 个人参赛记录不应因为注册账号就自动成为公开数据。
当前正式版本主要采用：
- Backend: Python / FastAPI
- Database: SQLite
- Frontend: HTML / CSS / JavaScript
- Deployment: Linux / Uvicorn / Nginx
线上版本还包含账号、数据维护、提醒、安全与部署等生产环境能力。
这个仓库是 Stellula 的公开项目展示与课程提交仓库。
它不是 Stellula 正式生产仓库，也不会作为线上版本的源码镜像。
目前这里主要用于：
- 介绍 Stellula 的产品目标与设计思路；
- 记录适合公开的项目说明；
- 作为 Python 课程设计的展示入口；
- 在课程提交阶段提供一份经过整理、可独立运行的 Mini Python 后端实现。
课程阶段会整理与 Python 学习目标直接相关的部分，例如：
- 基础赛事数据模型；
- FastAPI 路由；
- SQLite 数据访问；
- 专业关键词 / 映射逻辑；
- 竞赛搜索、筛选与排序；
- 适合课程演示的 API。
- Stellula 正式前端源代码；
- 生产环境完整后端实现；
- 真实用户数据；
- 管理后台与内部运营能力；
- OAuth、邮件、提醒等完整生产逻辑；
- 部署密钥、环境变量与安全配置；
- 内部数据维护流程及其他仅用于线上环境的实现。
答辩展示的是 Stellula；课程提交的是 Stellula Mini 的 Python 后端实现。
两者共享产品思想，但并不是同一份完整源码。
Stellula 仍在持续迭代。
这个公开仓库会保持克制：只有适合展示、学习和课程提交的内容才会进入这里。
Live: https://stellula.cloud
Stellula is a competition discovery platform designed for university students.
The problem is not that competition information does not exist. The problem is that it is scattered, inconsistent, and often difficult to evaluate.
A student may know their major and year of study and still have no clear answer to a simple question:
Which competitions actually make sense for me?
Stellula is built around the student journey:
Discover → Decide → Participate → Record → Look back
Instead of acting as a giant directory, it aims to help students narrow the field, understand why a competition is relevant, reach trustworthy official information, and preserve their own participation history.
Use information such as major, year, interests, and region to surface more relevant competitions.
Recommendations should not feel arbitrary. Stellula favors structured and understandable matching signals over opaque results.
Competition pages organize useful context and verified official resources. Stellula does not replace official registration channels; it helps students reach them with less friction.
Students can keep a long-term record of competitions they want to join, have joined, or have completed.
The wall is designed as a personal history, not a leaderboard.
- Useful before clever. AI is optional; the core product should work without it.
- Official sources first. Important information should remain traceable to authoritative sources.
- Explain recommendations. Students should understand why something appears.
- Designed for ordinary students. Prior competition knowledge should not be required.
- Participation is more than an award. Experiences deserve to be recorded without reducing people to rankings.
- Privacy by default. Personal competition history should remain private unless the student chooses otherwise.
The production application currently centers on:
- Backend: Python / FastAPI
- Database: SQLite
- Frontend: HTML / CSS / JavaScript
- Deployment: Linux / Uvicorn / Nginx
The production system also includes additional account, operations, reminder, security, and deployment capabilities.
This repository is the public project showcase and course-deliverable repository for Stellula.
It is not the production repository and is not intended to mirror the complete Stellula source code.
Its purposes are to:
- introduce the product and its design decisions;
- keep documentation suitable for public viewing;
- serve as a presentation entry point for a Python course project;
- later contain a small, independently runnable Mini backend implementation prepared specifically for coursework.
The course-oriented backend may include selected components such as:
- basic competition data models;
- FastAPI routes;
- SQLite data access;
- major keyword / mapping logic;
- search, filtering, and ranking;
- APIs suitable for classroom demonstration.
This repository will not publish:
- the production frontend source code;
- the complete production backend;
- real user data;
- internal administration and operations tooling;
- the full production OAuth, email, reminder, and security flows;
- deployment secrets, environment variables, or private infrastructure configuration;
- internal data-maintenance workflows and production-only implementation details.
The presentation uses Stellula. The course submission uses a Mini Python backend derived from the same product idea.
They share the same product direction, but they are intentionally not the same complete codebase.
Stellula is actively evolving.
This repository will stay intentionally small: only material suitable for public presentation, learning, and coursework will be added here.
Live product: https://stellula.cloud
Stellula
Small stars can still show you where to go next. ✦