Infrastructure for Personality-Aware AI
Why we built this.
What We Saw Was Possible
The conversations that actually move people forward — that help them think through hard decisions, challenge assumptions, discover what they believe — those conversations feel human because they adapt to how you think.
And here's what's interesting: how you think about climate policy is different from how you think about your finances. Your personality doesn't stay constant across domains. The most useful conversation is the one that sees that variance and works with it.
We're building the infrastructure to make that possible.
What We Believe
Personality isn't a label. It's how someone engages with ideas — and that shifts depending on the domain.
Conversations work better when agents understand not just who they're talking to, but how that person thinks in this specific context.
The right measurement method is observation, not self-report.
Technology should be honest about what it is and what it isn't.
Research credibility matters more than hype.
How We're Building
- 01 Start with measurement: can we reliably measure personality through conversation?
- 02 Validate the routing: does complementarity matching improve outcomes?
- 03 Build the infrastructure: a platform any AI system can use.
- 04 Publish what we find.
Where We Are Now
We're running a validation study.
- – Building CEP measurement infrastructure
- – Running structured debates to collect behavioral data
- – Testing whether personality-aware routing improves conversation quality
- – Publishing findings
Want to be part of this?
Participate