The archive · AI & Models · Product decision · 2024–2026
remio 2.0: ex-NetEase VP's local-first AI memory app tops Product Hunt
Ex-NetEase VP Wang Yuan's remio captures your whole digital trail locally and answers from it — #1 Product of the Day, 532 upvotes.
remio (Hangzhou Jiuhén Technology / 杭州久痕科技)
What the business is
remio is a local-first AI note-taker and personal knowledge hub that automatically captures web pages, documents, meeting recordings, chats and email on the user's Mac, then answers questions against that private index using DeepSeek or the user's own LLM API key.
Starting capital：Seed round in mid-2024 (amount and investors not officially disclosed); founder said a next round was in progress as of October 2025.
How it started
Wang Yuan spent nearly 18 years at NetEase — vice president and head of its Hangzhou Research Institute, helping incubate CloudMusic, Yanxuan and other products — and left in mid-2024 to found Hangzhou Jiuhén Technology. Having watched GPT-4 change what models could do through 2023, he concluded the next big opportunity was AI productivity tools for consumers, and that the market had to be global because Chinese tool ARPU was a fraction of Western levels. He chose local-first architecture partly to avoid colliding with the data-gathering business models of big tech.
What happened
remio opened a public beta in March 2025 as an AI note-taker that auto-captures browsing and builds a searchable personal knowledge base; the October 2025 version added folder auto-sync, saving of ChatGPT/Gemini/DeepSeek transcripts and YouTube subtitles, unlimited free recording transcription, email and Slack sync, and Markdown import. Wang positioned the product as 'context engineering for the individual' — an assistant that knows as much about the user's work as the user does. On January 14, 2026 the global remio 2.0 launch on Product Hunt finished #1 Product of the Day with 532 upvotes and 72 comments.
How it ended up
Still running as of mid-January 2026: app live with free tier, #1 on Product Hunt launch day, next funding round in progress per the founder.
Background
remio is a local-first AI assistant from Hangzhou Jiuhén Technology, founded in mid-2024 by Wang Yuan, a NetEase vice president for nearly 18 years. The app captures the user's digital trail — web pages visited, local documents, meeting recordings, chats and email — and indexes it entirely on the device. At query time it sends only small relevant context to an LLM such as DeepSeek, or to the user's own API key, positioning itself as a 'personal ChatGPT' that knows the user's work.
Wang's argument is that generic AI assistants fail because they lack context. He spent 2023 watching model capability advance, then chose consumer productivity for the AI era because productivity tools, unlike content platforms, win on value to the individual and have no network effects — so a small team can reach commercial close faster. He also chose global markets deliberately, estimating a Western user is worth roughly ten times a Chinese user for tools, and chose local-first storage to sidestep big-tech platforms whose models require data in the cloud.
After a public beta in March 2025 and feature releases through the year, remio ran its global remio 2.0 launch on Product Hunt on January 14, 2026. It finished #1 Product of the Day with 532 upvotes and 72 comments, was hunted by well-known community figure Chris Messina, and held a 5.0/5 average across 10 reviews. The founder's stated targets around the launch were tens of thousands of registered users, $1M ARR by early 2026 and $5M ARR in 2026.
What has to be true
- Capturing every information channel locally gives remio a complete personal context that cloud tools cannot legally or practically match.
- Local processing removes the data-upload objection that blocks privacy-sensitive professionals from trying AI assistants.
- Choosing consumers and global markets avoided China's low tool ARPU and the platform-giant competition Wang wanted no part of.
- No network effects meant the team had to find paying users fast, which pushed product-led launches like Product Hunt to the center of strategy.
What can be applied
Don't fight model makers on model quality; win on context. A personal AI assistant's moat is whose data it can see, and local-first processing turned a privacy stance into a structural cost advantage.
Aftermath
As of 2026-01-15, remio is live with a free tier and finished its remio 2.0 Product Hunt launch at #1 Product of the Day with 532 upvotes, 72 comments and a 5.0/5 rating across 10 reviews. Per the founder's October 2025 interviews, the roadmap was a Windows client (due late 2025), mobile access to the personal knowledge base, and productized to-do and reminder features built from meeting and email data; a next funding round was in progress. User and revenue figures after the launch were not yet independently reported at that date.
Sources
- remio 2.0 - Captures your work to build a searchable knowledge base | Product Hunt Launch Dashboard
- 对话汪源：AI时代个人的效率革命正在如何发生
- Expert Interview with Andrew Wang of remio
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card