How we source, test and stay independent

Sources

Every factual claim links to its source, and every source carries a label:

• Regulator A regulator’s own rule, bulletin or guidance, such as the NAIC.
• Our test Something we ran ourselves, published on Tests with its data and code.
• Independent reporting News outlets with editors, such as TechCrunch or The Register.
• Official docs TypeSafe’s documentation and the gateways’ own pages. Good for how things work.
• Vendor case study A customer story written by TypeSafe. Reported, not confirmed.
• Vendor claim TypeSafe’s marketing and blog. Reported, not confirmed.

Quotes are verbatim and checked against the page. Headlines named in our posts are real and linked. If we can’t confirm something, we cut it or say we couldn’t.

Testing

We test on public or synthetic data, compare against a baseline, publish the losing cases, and name the Jev version and date. See how we test.

Editing

Every post goes through research, a source check, a plain-language edit and a final human sign-off before it is published. Code samples are run, not just written.

Disclosures

Decision Layer is independent. It is not affiliated with, sponsored by or endorsed by TypeSafe AI. Jev is a product of TypeSafe AI.

Decision Layer is published by Conversely, which provides performance services to insurance agencies. [DISCLOSURE: Peter to confirm whether Conversely has any relationship with TypeSafe AI or a competitor, or builds with Jev for clients. Any such tie will be stated here and on every related page.]

No vendor pays for coverage, placement or reviews. We do not use affiliate links.

Not legal or compliance advice

Guides for insurance agencies describe operational tooling. They are not legal advice, and they do not replace the AI and insurance rules in your state.

Corrections

Found a mistake? Email info@converselyai.com. Corrections are listed here with the date and what changed.

No corrections yet.

Our sources on TypeSafe’s own approach

TypeSafe publishes a list of where its current model is weak (Jev 1.13 jaggedness) and argues against leaning on public benchmarks (Lies, Damned Lies, and Benchmarks). We hold its claims to the same standard.