Total insured value$185MAcross the insured portfolio
SAMPLE APPLICATION SCENARIO
Commercial Placement Review
Enterprise decision workspace for a complex, multi-location manufacturing renewal.
Insured locations4 statesThree high-priority sites
Placement strategies2 finalistsFollowing evidence review
Decision readinessHighKey evidence verified
EXECUTIVE OVERVIEW
In reviewRenewal strategy
Identify the strongest renewal structure without limiting the review to incumbent quotes. Validate material coverage gaps, market eligibility, and evidence capable of changing the final recommendation.
Lines in scopeProperty, GL, Auto, Umbrella
Primary exposureCAT and wind/hail
Open claims2 under active review
Risk improvementRoof remediation program
PLACEMENT PORTFOLIO
Updated 12 minutes agoStrategy comparison
| Strategy | Structure | Annual premium | Coverage assessment | Review status |
|---|---|---|---|---|
| Carrier D + CAT LayerRecommended structure | Primary + supplemental | $1.18M | Preferred | |
| Carrier CIncumbent alternative | Single carrier | $1.11M | Finalist | |
| Carrier DStandalone option | Single carrier | $1.02M | Gap identified | |
| Carrier ABroad incumbent quote | Single carrier | $1.24M | Not selected | |
| Carrier BAlternative quote | Single carrier | $1.06M | Not selected |
EVIDENCE REGISTER
View allMaterial findings
Market eligibility confirmed
Updated construction records opened an additional carrier option.
Roof remediation verified
Two sites completed; one location remains subject to restriction.
Supplemental CAT layer validated
Additional structure resolves the remaining protection gap.
GOVERNANCE
On trackReview readiness
Required evidenceComplete
Stakeholder review3 of 4
Decision briefReady
AM
Scenario ownerAlex Morgan · Risk Strategy
LIVE BENCHMARK — THE HEADQUARTERS DECISION
Same LLM. Same question.
POLOXI changed the ranking.
THE CHALLENGE PROMPT
I have $2 million to start a software and AI company in the United States in 2027. Which U.S. city should I choose for our headquarters overall?
We will start with 15 employees and may grow to 100–200 within five years. Consider software/AI talent, salaries and operating costs, housing affordability, ability to recruit and retain employees, access to investors, enterprise customers, universities, airport connectivity, taxes, quality of life, and long-term economic prospects.
We want to preserve our startup runway and ownership, but choosing a cheap city that limits our ability to recruit talent, reach customers, raise capital, or scale would also be a mistake.
Rank at least five serious candidates and determine the best overall choice. Do not treat all factors equally and do not assume the largest tech hub should win. Preserve uncertainty if the evidence does not clearly support one winner, explain the major trade-offs, and identify what factors could change the overall ranking.
| # | With POLOXI (evidence-weighted) | Evidence Score | LLM Raw (single-shot answer) | Raw pick's POLOXI rank |
|---|---|---|---|---|
| 1 | Austin, TX | 83% +3% vs next | Austin, TX | #1 |
| 2 | Raleigh-Durham, NC ▲ was #5 | 80% +6% vs next | Seattle, WA | #4 ↓ |
| 3 | Denver, CO ▲ was #4 | 74% | Boston, MA | #5 ↓ |
| 4 | Seattle, WA ▼ was #2 | 74% | Denver, CO | #3 ↑ |
| 5 | Boston, MA ▼ was #3 | 72% | Raleigh-Durham, NC | #2 ↑ |
SAME MODEL · SAME PROMPT · ONLY POLOXI'S EVIDENCE COMPETITION DIFFERS
The most interesting part isn't Austin. The raw LLM ranked Austin → Seattle → Boston → Denver → Raleigh. POLOXI produced Austin → Raleigh → Denver → Seattle → Boston—moving Raleigh from #5 to #2 and Boston from #3 to #5. That is exactly the behavior expected when an evidence-weighting mechanism is genuinely affecting candidate competition, not merely decorating the same answer.
RALEIGH
#5 → #2
Why it rose
Not strongest everywhere: investors/enterprise customers only 57%, AI talent 70%. But housing affordability 98% and operating costs 96%. For a startup with $2M of runway, POLOXI recognized those strengths deserve substantial influence—precisely the trade-off reasoning the prompt was designed to provoke.
BOSTON
#3 → #5
Why it fell
Spectacular numbers—AI talent 100%, investors 97%, long-term economics 89%—but operating costs 40% and housing affordability 26%. The prompt explicitly said “preserve startup runway and ownership.” POLOXI respected that constraint instead of concluding great talent ⇒ high rank.
SEATTLE
#2 → #4
The same pattern
AI talent 94% and investor access 89%, but operating costs 48% and housing 34%. POLOXI systematically moved expensive high-talent cities downward and capital-efficient balanced cities upward—a logically coherent transformation given the $2M constraint, not random rank movement.
The margins matter too. Austin beats Raleigh by only +3%, and Raleigh beats Denver by +6%. Austin isn't “the definitive answer”—it's a contested leader with exposed margins. POLOXI's next reasoning step can focus precisely on the factors separating Austin and Raleigh instead of re-researching all five candidates. This is what it means to demonstrate: here is what the exact same LLM chose without POLOXI, here is what POLOXI changed, and here is the evidence that caused the ranking to change.