How Neo answers
Neo is built to be trustworthy on questions where a wrong number is expensive. Four principles govern how it produces an answer.
Grounded in monitored data
Neo does not free-associate. Every answer is built from Bitpulse's monitored on-chain risk data and the same risk engine that powers Ozone. When Neo gives you a figure, it comes from a snapshot of real on-chain state, not from the model's general knowledge.
Every answer has a source
Each figure Neo reports traces back to the specific data it came from. The methodology behind a risk review is published — see Risk Review Methodology — so any number can be followed back to the check that produced it and the data that fed it.
Deterministic and reproducible
The risk review and the portfolio stress model are deterministic: the same inputs and the same snapshot produce the same output, every time. There is no run-to-run drift in the underlying analysis.
It abstains rather than guesses
When an input is stale, missing, or inconsistent, Neo says so instead of inventing a number. A missing estimate is shown as missing — never quietly rendered as low risk. Neo also surfaces data freshness, and works within a 90-day lookback window of recent on-chain history rather than extrapolating indefinitely.
If Neo declines to answer or flags a figure as "degraded," that's the safety model working. Ask what input it's waiting on, or narrow the question to data it can ground.
Deeper technical detail
This page describes Neo's trust posture, not its internals. For the portfolio tail model specifically, see Portfolio Stress methodology.