From AI recommendation to tested decision
A decision-validation layer powered by specialised AI agents.
Specialised perspectives
01 — Accio Work
Accio Work · external · opportunity discovery
Smart Water Bottle — Germany
Illustrative example
Accio Work is designed to identify a product, market or commercial opportunity using AI.
Market
Germany
Launch budget
€30,000
Target price
€129
Arbiter Trade · decision validation
Is this worth committing money to?
02 — Arbiter Trade
03 — Specialised AI agents
Critical Reviewer
Challenge the assumptions and identify contradictions.
In this method, agents analyse assumptions behind a single Arbiter Trade interface — rather than individual bots.
04 — What must be true?
Illustrative example
Critical assumption
“German customers will accept a €129 price.”
Recommended validation
Price test — 50 respondents
€300 tested before €30,000 committed.
Measured: declared purchase intent at €129. Threshold chosen for illustration.
05 — Decision record
Illustrative example
Timestamped before the test
Hypothesis: German customers will accept a €129 price.
62%
Expected acceptance
Stop threshold set in advance: below 45% acceptance, we don't launch.
Test result
36%
Outcome
Stop threshold reached: €29,700 not committed.
06 — Decision gate
Validated
Evidence supports the decision.
Evidence: All critical assumptions supported
Validate more
Important uncertainty remains.
Evidence: 1+ assumption still ambiguous
High risk
Critical assumptions remain unsupported or contradicted.
Evidence: Price assumption contradicted
In our example, the €129 price assumption leads to High risk.
07 — The system learns from outcomes
We want to publish how the system performs: what it anticipated correctly, where it was wrong, and where it is too early to conclude.
Company data stays private. Arbiter Trade's performance stays measurable.