Launching today
Pharen Hub
One AI workspace for your whole company
87 followers
One AI workspace for your whole company
87 followers
Pharen Hub brings projects, tasks, docs, mail, meetings, data and AI agents into one workspace with one shared context. Agents work with clear goals, allowed sources and human approval. Hosted in the EU or on your own servers. Fair-code, source on GitHub. Free plan to start.




















Ahoy, Product Hunt! I'm Johannes, founder of Pharen from Leipzig.
For a year we've watched teams drown in tools. Fifteen apps, a new AI tool every month, and people still copy-pasting between them all day. The AI could write, but it couldn't actually do the work, because it never had the full picture.
So we built Pharen Hub: one workspace where your tasks, knowledge, mail, meetings and AI agents share one context. Agents get clear goals, allowed sources and limits. Big decisions stay with people.
It runs in our EU cloud or on your own servers, and it’s Open Fair-Code. The full source code goes public on GitHub tomorrow, so you can read every line, self-host it and see exactly what the agents do.
Launch offer for this community: 10% off for 1 month with code LANDHO.
One question for you: which workflow would you hand to an agent first? We read every comment today.
@modiyilreji Thanks for your feedback!
Yes, we have an app as well which will be launched this month itself & you can handle approvals or tasks from anywhere with your phone!
Congrats on the launch! When the whole company shares one AI workspace, how do you decide what each person's AI is allowed to see? That's the question we hear most from teams rolling out AI internally.
@tsubasatakeuhi Thank you! That’s the right question, and we hear it a lot too. In Pharen Hub every piece of data has its own visibility setting: each doc, list, recording, agent and workflow. On top of that there are workspace and team settings, so you decide exactly who sees what. An agent only works with what it’s allowed to see, and the sources, tools and approval steps are set per agent. Every run is logged as well, so you can always trace what the AI used and why. We’d rather teams start narrow and open up step by step than give the AI access to everything on day one.
@johannes_kanthak makes sense, thanks Johannes. "start narrow and open up step by step" is pretty much what we tell teams too
@tsubasatakeuhi Happy to connect on LinkedIn if you want more info!☺️
@johannes_kanthak sounds good, thanks! 🙌
Giving agents allowed sources as well as a goal is a useful detail. I'd want an approval to show which source and change it relates to, so a person can check the decision without opening every tool again.
@oleksii_sekundant Thanks, that’s exactly how it works in Pharen Hub. When an agent asks for approval, you see what it wants to do, which source it used and what will change, right there in the approval. You can check and decide without opening any other tool, and if something is unclear, the agent flags it instead of guessing. Every run is also logged, so you can always look back at why something happened.