We had 23 productivity apps open at once. So we built a second brain instead.
We didn't set out to build a productivity tool.
Six months in, our team had a tab problem. Notion for docs. Evernote for old notes. Google Keep for quick captures. Obsidian for the second-brain experiment that never quite worked. Todoist for tasks. Three tabs of ChatGPT, Claude, and Gemini open at all times. A shared Drive folder that nobody could find anything in. A bookmark list we'd given up on years ago.
Everyone on the team was using a different stack. Everyone was paying for five subscriptions. Nobody could find anything.
That's when we started building Brainfo.
The core idea is embarrassingly simple: one workspace where your notes, docs, bookmarks, databases, tasks, and goals actually live together — and a single AI hub that lets you switch between every major model (ChatGPT, Claude, Gemini, DeepSeek, Grok, GLM) on one subscription.
The harder idea is what NOT to build.
We're not trying to be another "all-in-one" workspace that demands a weekend of template setup before you can take a note. We're not trying to be Notion-with-AI. We're not building an outliner. The thing we kept coming back to was: it should feel like a calm place. Powerful enough for serious work. Simple enough that you don't have to "learn" it.
The trade-off most second-brain tools force you to make is real:
- Notion is powerful but you spend hours building the system before you do any work
- Obsidian is private and powerful but you need plugins and a weekend of setup
- Evernote is familiar but flat — no links, no databases, single AI model, getting expensive
- Google Keep is simple but you outgrow it in a month
- Tana is AI-native but the supertag learning curve is brutal
We're trying to land in the middle: ready out of the box, AI-native by default, with deep tools (real databases, real docs, real collaboration) that don't require system-thinking to use.
Some honest things we're still figuring out:
- How much pre-built structure is too much? Wrong answer kills the "make it your own" feeling
- Multi-model AI on one flat subscription is hard pricing math — but it's the whole point
- Local-first plus real-time collaboration is technically annoying. Worth it
- We're competing against the entrenched habit of "I already have my system"
We're currently in early access at brainfo.ai with a growing waitlist of students, writers, creators, freelancers, and small teams. If you've ever closed all your productivity tabs and thought "there has to be a better way" — I'd love your feedback. And if you've tried building in this space and gave up, even more interested in what made you walk away.
What are you using as your second brain right now? Curious whether anyone has actually solved the "I need it powerful but I also need it usable" problem in a way that stuck.
This resonates a lot.
I've tried a bunch of second-brain tools over the years, and the biggest problem always ends up being the same: the more tools you add to stay organized, the more time you spend managing the tools themselves.
The tricky part doesn't seem to be capturing information anymore. It's knowing what actually matters later and turning it into action without losing context.
Curious, have you found one specific reason people switch to Brainfo? Is it mainly to reduce app overload, or is there another pain point that's driving adoption?
Most second-brain setups slowly drift into tool overload because each layer solves one part but never unifies the actual workflow.
The harder part usually isn’t storing information, it’s deciding what becomes actionable later without losing context.
When people switch tools, what’s the first thing they tend to migrate — notes, tasks, or AI chats?
The thing I'd be careful with is that almost every tool you mentioned is solving a slightly different job.
The risk isn't having too little structure or too many AI models.
The risk is that Brainfo ends up competing with six different products before users are sure what problem they're hiring it to solve.
That sounds subtle, but it tends to shape everything that comes after.
I wouldn't make that call casually in a thread.