What competitors filed.
And what happened next.
Every U.S. rate filing discloses part of its actuarial justification. Reading competitors' filings by hand costs hours per question, so most of those questions go unasked. unifi.ai reads the public record for you — and joins it to the loss ratios that followed.
Access is invite-gated while we onboard the first cohort.
Three things a pricing actuary can do on Monday morning
Rate actions vs. results
A competitor’s approved rate changes beside the loss ratio that actually followed — alone, against a peer group you choose, and against every published writer. No filing-data subscription joins those two sides.
Rating variables across carriers
One variable, many carriers. Deductible relativities rebased onto a common base, because they are not comparable as printed. Selected loss trends per coverage, checked against the frequency × severity identity.
Filing reader
Upload filings from any state and get the elements of the actuarial justification out — with the page they came from, and honest blanks where the public copy does not disclose a value.
What it does not do
Stated up front, because an actuary will find these out anyway and would rather hear them from us.
- Loss ratios are direct, not combined
- The free public source carries no expense data, so a combined ratio is out of reach. We do not estimate one.
- Only the largest writers are published
- Realized loss ratios cover the top ten writers per state and line, so roughly a third of filings find a match. The long tail is invisible.
- Trend and development are often withheld
- Selected trends appear in about a fifth of filings, development factors in a seventh. That is a ceiling of the public record, not something better tooling lifts.
- Not a comparative rater
- We do not reconstruct a competitor’s rating algorithm, cannot price a given risk, and hold no quote or shopping data.