📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 43 分 ← 返回列表

I got tired of researching through endless AI chats and browser tabs, so I built an app that turns research into a visual map (Promo codes!)

Hey everyone! I've been working on a side project called Branchscape , and I finally feel like it's ready to share. The idea came from something that kept annoying me whenever I was researching anything online. I'd start with a simple question, either in ChatGPT or Google, get a few interesting answers, and naturally want to explore some of them further. But after 20 minutes, I'd either have 15 browser tabs open or a ridiculously long chatbot conversation where I had to

五维评分

付费意愿
30
需求真实度
35
个人可落地
82
竞争空间
28
新鲜度
45

AI 分析

痛点:做 AI 或网页调研时,信息分散在大量浏览器标签页和超长聊天记录里,调研路径并非直线,因此难以回溯各条线索之间的关联,最终丢失上下文

目标用户:需要频繁做多线索深度调研的知识工作者、学生、分析师、内容创作者

证据:after 20 minutes, I'd either have 15 browser tabs open or a ridiculously long chatbot conversation where I had to keep scrolling up to find something I'd read earlier... I'd discover three interesting things, want to explore each of them separately, and eventually lose track of how everything connected.

变现:Freemium 订阅:免费额度限制每月研究分支数量,付费解锁无限分支、导出(Markdown、PDF、Notion)、团队共享与研究历史归档;也可做 API 或白标嵌入

MVP:用 Next.js 加 LLM API(进行 web search 和结构化抽取),把每个问题渲染成卡片节点,用 React Flow 或 D3 做可拖拽的分支画布;支持点击卡片上的 + 号追问并生成子节点,再支持导出 Markdown,两周内可上线

竞品:Perplexity、ChatGPT Deep Research、Heptabase、Scrintal、Obsidian Canvas、Napkin AI

摘要:作者把“调研时开15个标签页、聊天记录越滚越长、线索间的关联难以看清”这个痛点做成了 Branchscape:AI 联网调研后以可视化卡片和可分支节点呈现结果。帖子属于自我推广的 launch 内容,没有付费意愿、收入或第三方验证的数据,且该方向已有 Perplexity、Heptabase 等成熟竞品,属于需求真实但变现证据薄弱的红海方向。