Replies: 11 comments 8 replies
-
全局异常处理 |
Beta Was this translation helpful? Give feedback.
-
所有面评记录完成后,可以通过 ab 内部 gpt sdk 来快速对所有面评打分评价确定不同的类别, ab 内部 gpt java sdk: AchoBeta-OpenAI-SDK |
Beta Was this translation helpful? Give feedback.
-
短链生成 |
Beta Was this translation helpful? Give feedback.
-
数据库表设计:面试时间预约功能 |
Beta Was this translation helpful? Give feedback.
-
管理员设置面试时间后,通过腾讯会议 API 新增会议链接并存储到招新系统后台 |
Beta Was this translation helpful? Give feedback.
-
无密码登录相关设计文章: |
Beta Was this translation helpful? Give feedback.
-
token 进阶:无感刷新 token |
Beta Was this translation helpful? Give feedback.
-
【扩展】: 面试题库功能(可结合 GPT 生成),通过对简历内容进行分析,搜罗面试问题清单。(这个功能是否要做有待商榷,也可在本评论下回复讨论可行性) |
Beta Was this translation helpful? Give feedback.
-
面试日程提醒功能: |
Beta Was this translation helpful? Give feedback.
-
功能可参考华工华南虎招新系统,但得有一定创新 |
Beta Was this translation helpful? Give feedback.
-
组卷系统 一来极大程度减少重复开发(后续如果真的有必要拆再拆),二来能有效落实面评的记录,毕竟已经细化到问题粒度,而不是对面试者的总体评价了,而总体评价则变成了各个问题的加权平均以及面试官对面试者的印象分 面试类问题可以搞个排名机制,被抽到的频率有多高,谁答得最好,通过率多少(难易程度),都可以做统计,这样可以实现梯度化难易更替,比如大一对于计网类问题普遍回答都差,那就降低权重,减少组卷系统对其的选择;这样一来,经过几次招新就能知道大一大二的平均水平 |
Beta Was this translation helpful? Give feedback.
-
总纲:接口一定要加密处理,防止被反爬
架构选型
额外:
Beta Was this translation helpful? Give feedback.
All reactions