Drop in a carrier loss run. LossLens reads every claim the way an underwriter would — totals, trends, the risks worth flagging, a renewal-ready write-up — then independently re-checks every figure against the source before you rely on it.
Thirty to sixty minutes of reading and re-keying per account. One transposed reserve and the whole risk picture is wrong. A generic AI tool will hand you tidy numbers and hope they’re right. LossLens does the reading and proves the numbers tie out.
Upload the carrier PDF — several policy years or carriers at once — or paste the text. Scanned runs work too.
Every claim, the totals and year-over-year trend, the risk flags worth raising, and a tight renewal-ready narrative. In seconds.
Every figure is re-added in plain arithmetic and reconciled to the total printed on your run. Green means it ties out; amber points you to the one number to check.
Most AI tools give you an answer and move on. LossLens re-derives every total straight from the claim rows — no model involved in that step — and cross-checks it against the figure printed on your carrier’s run. If anything doesn’t reconcile, it flags the exact number to look at.
You never bind on an unverified figure — and every report still says to confirm against the carrier run before you underwrite. How we handle your data →
Incurred, paid, open reserves, claim counts, and the year-over-year severity trend.
Severity concentration, open development, frequency trends, recurring loss types — quantified.
Every claim, with a reference back to where it appears in the document.
A neutral, editable paragraph you can drop straight into a submission file.
Send an underwriter a clean, login-free view, or export a PDF with your agency’s name.
A dated, referenceable receipt (LL-XXXX) of what reconciled — it ships in the export.
The PDF is processed to produce the analysis and not written to our database — only the structured results are saved to your account.
Analysis runs on Anthropic’s Claude API under commercial terms, which state inputs and outputs aren’t used to train their models.
We don’t yet hold SOC 2. The Security page lays out exactly how data is handled — no hand-waving.
LossLens doesn’t make coverage, admissibility, or compliance calls. You and your carriers make every decision.
I’m signing a handful of agencies at founder pricing in exchange for shaping what gets built and a direct line to me. Month-to-month, no long contract. Tell me the workflow that eats the most time and we’ll start there.
A synthetic sample — no signup, no real client data. Takes about ten seconds.
Run the sample analysis →