OpenHuman
An open source AI harness built with the human in mind
1.9K followers
An open source AI harness built with the human in mind
1.9K followers
90% of people who try AI agents give up. Three reasons: memory that resets every session, your data sitting in someone else's cloud and a terminal just to get started. Real blockers. OpenHuman fixes all of it. Local-first, privacy-first. It remembers everything about you and actually gets smarter the more you use it. Every feature lives in one simple interface. Fully open source. One-click setup. P.S. The product is in beta, so expect bugs, but we're building and shipping fast.
The dad test is a good framing and a real problem. The 90% dropout stat is doing a lot of work though. Where does that come from?
The memory claim is the one I'd want to stress-tested. "Remembers everything about you and gets smarter" is a strong promise. What's the actual architecture: vector store, fine-tuning on local data, something else? And what happens to that memory when the app updates or the local storage gets corrupted?
8000 GitHub stars in early days is a real signal. What's the retention curve look like past day 7? That's where local-first tools usually fall apart.
How it's handle hallucination thing?
OpenHuman
@imrulkaayes Good question.
Three layers here:
First, OpenHuman grounds its answers in your actual data (emails, slack, notion, etc.) rather than generating from training memory alone.
Second, every memory chunk has a deterministic ID and we can show you the exact source for any claim the agent makes.
Third, when the agent isn't confident, it tells you and asks rather than guessing. We're not pretending hallucinations are solved, but grounding in your real corpus plus auditable retrieval cuts the worst of it.
Happy to go deeper if useful.
OpenHuman
@imrulkaayes The default prompts used by the agent already prevent halucination. And if you want to use a model that is completly yours and high end, you can easily switch to that as well. Most frontier models if prompted well do not halcuinate (much).
Lancepilot
OpenHuman
@odeth_negapatan1 thank you! 🙏 most surprising use case so far a florist in ohio using it to auto-reply to wedding inquiries on emails, manage her calendar, and send invoice reminders. zero coding background. the world is your oyster! that's exactly why we built it. agents shouldn't just be for devs.
OpenHuman
@odeth_negapatan1 Someone is using OH to unsubscribe from all the shitty newsletters and manage their instagram and twitter to grow their businesss. More incoming.
And super awesome to meet ya :D thanks and i hope you get to try out the product and give us more feedback
OpenHuman
One of the weirdest things about using OpenHuman now is how fast normal AI tools start feeling “dumb”.
The moment your AI starts remembering context across chats, understanding your workflow, pulling information across apps, and quietly helping in the background… it’s really hard to go back.
Feels less like using ChatGPT and more like having a second operating system running alongside you.
Also watching non-technical people set this up in like 2 minutes after spending months hearing “AI agents are too complicated” has been pretty wild !
OpenHuman
@kunal_karani welcome to the journey brother
OpenHuman
@kunal_karani welcome to the fam guy!
OpenHuman
Hi 👋
I'm Ankita, the product marketer at OpenHuman.
I joined this team because Steven and the rest of TinyHumans were the only people I'd met who genuinely wanted to build an AI agent for everyone, not just engineers. The people who don't write code, don't want to wire API keys, don't want to read YAML files. People like my parents, honestly.
Watching the last few months of shipping has been wild. The skills marketplace went from a handful of integrations to 118+. Memory went from session-only to actually remembering you across weeks. And Tiny the mascot has somehow become the most-discussed feature in our internal slack.
If you've ever wanted to use an AI agent but felt the setup wasn't worth the headache, today is your day. It gets way more useful the longer you use it because it actually remembers what you tell it.
Try it free at tinyhumans.ai/openhuman.
Hop into our Discord if you want to chat, ask for a feature, or report a bug. I'm reading everything today.
Thanks for being here 🙏
Ankita
OpenHuman
@erdoan1083411 Imagine it's like your own personal assistant running 24x7. OpenHuman is supposed to feel "Human". That's what special about it.
OpenHuman
@naciyeodef27994 You can actually speak with the mascot and it feels really human. This is one of my most favorite features and it comes built in with OH.