📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 53 分 ← 返回列表

I built an AI tool that ranks live job postings by how well you actually fit them

I got tired of watching people apply to 200 jobs blind, so I built FitScout on nights and weekends (React, Supabase, n8n). Upload a PDF resume, pick a role and location, and it scores you 0–100 against live postings in ~2 minutes, with a plain-language reason for each score. Weighting: skills 50%, responsibilities 25%, seniority 15%, location/type 10%. Just shipped v2: full redesign, faster parallel scoring, on-site + hybrid roles (searches were accidentally remote-only before), and a fix

五维评分

付费意愿
48
需求真实度
45
个人可落地
88
竞争空间
33
新鲜度
45

AI 分析

痛点:求职者常常不加筛选地投递上百份简历,却无法快速判断哪些在招职位真正匹配自己的背景,因此浪费大量时间。

目标用户:正在找工作的求职者,尤其是会不加筛选地大量投递简历的中初级白领、技术岗位求职者,以及需要快速校准投递方向的转行者。

证据:I got tired of watching people apply to 200 jobs blind / The free run scans 30 roles. Paid runs are $1.99 CAD, no subscription. / do the top matches actually feel right for your background, or do the scores seem off?

变现:按次付费解锁完整扫描报告(例如 $1.99/次),用免费额度获客;再叠加简历关键词优化、定制求职信、面试准备包等增值付费项,后期可做月度订阅包。

MVP:2 周 MVP:支持上传 PDF 简历并抽取技能;接入 JSearch/Adzuna 等职位 API 拉取实时岗位;用 LLM 按技能 50%、职责 25%、资历 15%、地点类型 10% 加权打分,并给出自然语言理由;免费送 30 条结果,付费解锁全部结果和排序导出。

竞品:Jobscan、Teal、Simplify、LazyApply、Rezi

摘要:独立开发者用 React + Supabase + n8n 做了 FitScout:上传 PDF 简历后,2 分钟内对实时职位给出 0-100 匹配分(技能50%、职责25%、资历15%、地点类型10%),并附评分理由。免费可扫描 30 个岗位,付费为 $1.99 加元单次、无订阅,v2 刚上线并修复了搜索结果只涵盖远程岗位的问题。帖子只体现定价设想,未披露任何付费转化或收入数据,付费意愿仍需验证。