Tesora

Tesora is AI native by its own description, built as agents rather than as software with models added. Strip the models and there is no workbench left, only the spreadsheets and email the product exists to replace. Where larger pricing vendors keep a deterministic rating engine underneath their models, Tesora never had a rating engine to begin with.

The division of labor is stated explicitly rather than implied. The product handles the assembly and the actuary stays in charge of every decision, a positional claim about where the human sits rather than a general assurance that humans are involved. Reviews are described as aligned to actuarial professional standards. That names an external standard the review is measured against, the same kind of specificity that separates an auditable control from guardrail language. Two limits remain.

Alignment to a standard is not enforcement of it. Nothing describes what a review actually checks, what an agent may do without approval, or what prevents an agent from deploying a rater the actuary has not examined. The agents do act, since building and deploying a rater is acting.

The evidence rests on two named mechanisms rather than assertions, and the vendor frames the problem correctly. It states that generic AI was untrustworthy because its analyses were often unreproducible. Its answer is reproducibility as a product property, plus provenance. The agents record where data and logic came from so they can cite the source later, and rebuilt raters become versioned models. Reproducibility and source citation are exactly what a reviewing actuary or examiner needs.

Much is missing. No accuracy figures, validation methodology or drift monitoring are published, and there is no error rate on the document extraction the whole pipeline depends on and no external assessment. The most checkable claim on the site is that when an existing spreadsheet rater is rebuilt, existing logic is preserved exactly. That is an absolute stated about a translation process, with testing mentioned and no test method, tolerance or failure rate published.

There is no named customer, quoted executive or case study, and no quantified outcome or independent evaluation. Design partners are referred to and never identified. Two things exist that are not customer evidence. An accelerator backing is investor validation, and a former chief actuary working with the team is a credential of the vendor rather than a reference from a buyer.

Nothing states whether a carrier's loss experience, rating logic or submissions inform anything beyond that carrier's own work. The question matters because rating logic is among the most commercially sensitive material an insurer holds, and the agents ingest it directly. The same agents also read competitors' public rate filings, so the product sits between public and proprietary rating information by design.

Nothing published on retention, access or processing commitments, for a product whose inputs are a carrier's loss runs and broker submissions containing insured and claim level detail.

The only security claim invokes a statute nobody certifies against. No certification, security page or trust portal is published. The audit trail is described as meeting the standard expected for financial reporting controls, which invokes the Sarbanes Oxley Act (SOX) as a quality standard. That statute governs the filer's controls rather than a supplier's product.

Statutes and uncertifiable frameworks are sometimes listed alongside real certifications, and this is a milder version of that pattern. The claim does credential work while naming no assessor, scope or report.

As a software supplier at this stage, Tesora claims no license or standing, and none would be expected. The product reads state rate filings as an input. That is consuming public regulatory data, which is a different thing entirely from having a model examined and approved by a regulator.

The rating agents build the price a consumer is charged, and no fairness testing, protected characteristic handling or governance disclosure is published. The exposure is direct rather than incidental. The agents produce base rates, factors and increased limits factors, which is the construction of that price. At least two competitors disclose bias testing, while Tesora, whose agents build the rating structure itself, does not.

Nothing published on responsibility for a wrong outcome, and the exposure is concrete. The agents build and deploy raters, so an error does not stay in an analysis. It reaches the market as a rate charged to policyholders. The spreadsheet rebuild path is the sharpest version.

A translation error in parsing formulas or identifying factor tables produces a rater that prices differently from the one the carrier approved, and nothing addresses detection, correction, or who bears the cost of business written on it. Reproducibility means the error can be traced afterwards, which answers a different question from who is responsible for it.

Tesora names no model, provider, family or version. The capability is described as frontier agents, a phrase that gestures at model quality while disclosing nothing about the dependency. Like "proprietary models", it is a term that substitutes for a name. The product's whole value proposition is that generic AI could not be trusted for this work. The claim that this system is different therefore rests on the model layer, and the model layer is unnamed.

Two real capabilities and no named integrations. The product reads state rate filings, loss runs, statements of value and broker submissions in almost any file format, and it exposes a callable pricing interface so a carrier can invoke a rate from its own systems, which is a genuine production path rather than an export.

Against that: no policy administration system, rating platform or data provider is named anywhere, no interface documentation is published, and the deployment target for a rebuilt rater is the vendor's own runtime rather than a carrier's existing stack. Ingesting documents is not the same as integrating with the systems of record.

Nothing published. No hosting model, no cloud named, no options, no residency statement, no customer responsibility split, for a product that ingests a carrier's loss experience and broker submissions.

No pricing, no commercial model and no indication of whether the product is licensed by seat, by rater, by call volume or by subscription. Every route is a contact form.

Coverage is a stated target, not yet a customer base. Founded in 2025 with a founding team of two to four, the company describes working with design partners. Target buyers are stated with precision: specialty property and casualty carriers, reinsurers and managing general agents. No institution of any size is named and no count is claimed. This is a startup at the beginning of its commercial life rather than a vendor failing to disclose.