What AUR is
A research lab built around a simple public question: can a swarm of AI agents discover a real, measurable market edge without fooling itself first? AUR is not a fund, broker, or signal service.
Independent build-in-public market research
Can a swarm of AI agents discover a real, measurable market edge without fooling itself first?
AUR Research Lab studies that question with a bias toward falsification, careful timing, realistic constraints, and public-safe evidence. The aim is not to make a promising backtest sound certain. The aim is to learn what survives honest scrutiny.
01 / The lab
Research before confidence.
AUR is an independent project exploring whether AI-assisted research systems can find useful, repeatable market structure without mistaking noise, leakage, or convenient stories for evidence.
What it studies
The work looks at market microstructure, liquidity, forced flows, cross-venue behavior, data quality, and the design of falsifiable AI-assisted research systems.
What it protects
Every result must earn trust through causal timing, falsification, out-of-sample or prospective validation, realistic costs, and an audit trail that makes failure visible.
02 / Build in public
Share the reasoning. Protect the work that is not ready to share.
Build in public means making the questions, standards, lessons, and safe decisions legible while keeping active signals, sensitive parameters, private infrastructure, and unvalidated results out of the public record.
No hype. AUR does not currently sell signal access, investment advice, subscriptions, or a public performance-fee product.
No guaranteed returns. AUR does not publicly claim a durable edge, and this site does not publish performance claims, predictive results, or active implementation details.
03 / Current public stage
Disciplined exploration, in the open.
The public stage is about establishing a credible research habit: define questions clearly, preserve an audit trail, document what fails, and publish only lessons that are useful without exposing sensitive research mechanics.