📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 46 分 ← 返回列表

I got tired of big tech scanning files and using calls to train AI, so I built Anchor Cloud: Zero-knowledge storage with E2EE video meetings built in.

Hey everyone, Like many of you, I’ve grown increasingly uncomfortable watching major cloud providers update their terms of service to scan uploaded documents, read chat attachments, and harvest video call data for "AI product improvements." The current privacy ecosystem is fragmented: if I want to have a confidential consultation with someone, I have to hop on Signal or an E2EE video room, but then jump over to a separate encrypted storage provider just to share documents. The moment some

五维评分

付费意愿
40
需求真实度
45
个人可落地
55
竞争空间
35
新鲜度
60

AI 分析

痛点:在隐私敏感的专业沟通中,加密存储和端到端加密(E2EE)视频会议彼此割裂;只要把敏感 PDF 拖进 Zoom 或 Teams 聊天,零知识保证就会立即失效。

目标用户:律师、心理咨询师、财务顾问、记者等需要保密会谈与小团队协作的从业者

证据:“The moment someone drops a sensitive PDF into a Zoom or Teams chat, the zero-knowledge guarantee is broken because the meeting provider stores that attachment in the clear.” 以及 “I spent the last few months building Anchor Cloud”;全文没有付费用户、收入或第三方附和的证据。

变现:按席位订阅(专业版 8–15 美元/席位/月);合规增值包(HIPAA/GDPR、数据驻留、审计日志、会后自动销毁);向小型律所、诊所出售自托管授权。

MVP:两周内只做“加密房间 + 加密文件投递”闭环:浏览器端用 WebCrypto AES-GCM 加密文件并生成一次性密钥链接,会议采用现成 LiveKit/Jitsi 的 E2EE 房间,不做完整云盘和跨平台客户端。

竞品:Proton Drive / Proton Meet、Tresorit、Internxt、Sync.com、Element Call (Matrix)、Signal、Filen、Cryptomator

摘要:独立开发者因不满大厂扫描上传文件、用通话数据训练 AI,用几个月做出 Anchor Cloud:客户端 AES-256-GCM 零知识存储、自托管 LiveKit 帧级 E2EE 视频会议,以及通话内加密文件投递。帖子属于产品发布,未披露任何付费用户或收入数据,需求目前主要是作者的个人痛点。可借鉴的切入点是“加密沟通 + 加密文件”一体化工作流,但该赛道已有 Proton、Tresorit、Element Call 等成熟玩家,竞争空间有限。