A calibration study to validate the software on real worksheets.
Premwise runs a calibration study in select states: we run our recomputation software against real experience-rating worksheets to measure how often a bureau’s published mod actually reproduces from its own inputs. It’s research that validates the software — not a sales program.
Why yours might be worth checking
Since 2024, every state uses its own split point in the mod formula — and the values change with every annual filing. The mod is built from claim and payroll data that passes through three parties before it reaches you. An EMR affects premium and may be one requirement in a prequalification process, so reproducing its arithmetic matters; the project owner, general contractor, or prequalification service sets any threshold and other requirements and determines eligibility.
Every state’s current filed values are published in our free lookup, with sources and effective dates — they change with every annual filing, which is part of the point. See the filed state values →
The study runs in select states; eligibility for your state is confirmed when you’re invited — before anything starts.
What you get back
- The full recomputation, in a single document titled “Experimental recomputation result”: every value we read from your worksheet and the arithmetic performed on it.
- The filed state values the software used — each with its public source and effective date, so you (or your broker) can check every number by hand.
- Whether the printed mod reproduces from its own inputs — including exactly how far off it is if it doesn’t.
No fee, now or later. The study calibrates the software; honest results — including “your worksheet reproduces exactly” — are exactly what we need.
What it takes on your side
Two signed study documents and one file:
- The Software Calibration Study Agreement and the Study Authorization — both presented in full, in the app, before anything starts. You sign the documents themselves, not a summary.
- One document: a copy of the experience-rating worksheet your company already holds.
That’s the whole ask. Premwise obtains nothing from any carrier, rating bureau, or third party — the worksheet comes from you, and the study touches nothing else.
Premwise never contacts your carrier or the rating bureau — for the study or anything else. The result is experimental and not a submission-ready document; what you do with it, you decide with your own licensed broker or counsel.
The authorization’s operative scope, verbatim
Participants sign two instruments in the app — the Study Agreement and the Study Authorization — presented whole before anything starts. This is the Authorization’s operative scope paragraph and the acknowledgments you’d sign:
Software Calibration Study Authorization
This authorization permits Premwise to process only the worksheet that Participant already possesses and uploads for the free software-calibration study. It does not authorize Premwise to obtain records from, contact, submit to, negotiate with, or represent Participant before any carrier, rating bureau, regulator, broker, project owner, or other third party.
Signer acknowledgments (each required):
- I am authorized to sign for Participant and provide the worksheet.
- Participant already possesses the worksheet and will upload it directly.
- I understand that Premwise is not authorized to obtain records from or contact a carrier, rating bureau, regulator, broker, or other third party.
- I understand that the study provides only an experimental recomputation result and no recommendation, correction package, submission, negotiation, statutory analysis, or recovery service.
- I understand that Participant owes no fee and that a study finding cannot become the basis of a later Premwise recovery fee.
- I have reviewed the Software Calibration Study Agreement and agree that the more restrictive limitation controls if the documents conflict.
How your data is handled
- Used only for the study, system security, legal compliance, and the permitted output.
- Never sold — and never used to train or improve a generalized or shared machine-learning model.
- The raw worksheet is deleted from active systems within 90 days of study completion or withdrawal.
- Revocable — written notice to support@premwise.ai stops new processing; revocation does not undo processing completed before it became effective.
Who’s behind this
Premwise is software that recomputes experience mods from the rating bureaus’ own filed rules and values, tested against their published worked examples. This study is how it meets real worksheets. We don’t sell insurance, take commissions, or recommend carriers — the study’s only purpose is to validate the software and, for participants, to hand back a careful, cited check.
Interested in taking part?
Email support@premwise.aiEmail us and we’ll explain what participation involves — including the Study Agreement and the Study Authorization you’d review in full before anything starts.
Premwise · A software calibration study — not legal, insurance, or financial advice. Results are experimental; rating bureaus and carriers make their own determinations.