Data research lab for the non-verifiable
Hello. We are Lance Yan and Zachary Yu, the founders of Traverse.
TLDR: Traverse is a data research lab focused on the hardest unsolved problem in frontier model training: ambiguous, taste-dependent work.
We capture real human behavior in highly controlled settings and turn judgment, taste, and long-horizon reasoning into high-fidelity environments for frontier models.
The Problem
Reinforcement learning has already helped frontier models achieve superhuman ability in math and coding since these domains have very clean and deterministic feedback loops.
However, the next frontier domains (law, healthcare, sales, writing, etc) aren’t as black and white. Current approaches fail to accurately represent subjective scenarios, despite these domains being where most of the economic value is.
The Solution
We are building data factories to create high fidelity real life work environments at scale.
Traverse captures what people actually think while doing their work; we are pioneering a new category of training data that is yet to exist.
If you’d like to understand the details, reach us at
founders@traverse.so
https://traverse.so
