Fair question
Why not just use ChatGPT?
Open Nexus runs on Claude, a model made by Anthropic, so the language skills come from the same kind of technology. What's different is everything we built around it.
A real example
"I've been working twelve-hour shifts. I love my job, but lately I don't even want to get out of bed."
Both sides of that sentence show up in the read: the weight (sadness, 41%) and the part that still loves the job (joy, 24%), with some fear underneath.
And it tells you how sure it is: signal agreement, 95%.
It shows its work
The Observatory lays out what it picked up in your words and how sure it was, so you can check the read instead of just trusting it.
It's built for people who understate
People trained to stay calm write "rough shift" when they mean a lot more. Open Nexus is tuned to notice that. Patent pending.
Nothing goes to your workplace
Not your words, not your name, not how much you use it. If your department runs CHR, only a group number crosses, and only for groups of eight or more.
Your words don't stay forever
What you write is destroyed after 90 days, and you can erase all of it any time, without telling anyone why.
Your messages are sent to Anthropic so the AI can reply, and they aren't used to train their models. Every detail is on our Privacy page.


