jev.proField guide · curious → Pro

Open · Last updated

Jev testimonials

Jev testimonials on jev.pro are public, named quotes already printed in the press — framed as “what builders are saying in public,” not as a wall of five-star cards we collected. Jev is TypeSafe’s System One decision model. We do not invent customers, stars, or “Acme Corp saved 80%.”

This page is for people who searched testimonials / social proof and should leave with source links, not a landing-page collage. Context and jagged edges: Jev review. Company: TypeSafe AI.

How to read these

Builders in TechCrunch (18 September 2026)

Tim Fernholz, TechCrunch:

Pranit Sharma, software engineer at Vercel — paraphrase in the article, not a verbatim quote. TechCrunch writes that Vercel had used OpenAI’s ChatGPT Luna 5.6 to classify commands for safety; when they replaced Luna with Jev, “it got results five to 18 times more quickly and with greater accuracy.” Those multiples are Sharma-via-TechCrunch, not a jev.pro benchmark.

Nikhil Mudholkar, CTO of Bryo AI — tested Jev against Gemini for classifying business emails. TechCrunch: Gemini was slightly more accurate in his test, but 10 to 20 times more expensive. Direct quote on confidence scores:

“it is the only one that hands back a real probability which makes it ideal for automating workflows!!”

Armin Ronacher, CTO of Earendil (Pi harness). On using the probability instead of treating the model as an oracle:

“At the end of the day, it delegates the hallucination problem a little bit to the user. The user has to say, okay, if this only comes back with 50% probability, maybe this is a coin toss, and I disregard it. But if it’s 95%, sure, then I can do something with it.”

On why this shape of model showed up late:

“We should have seen this earlier in many ways, but presumably because the LLMs are so cheap and subsidized, you often don’t have to be creative yet.”

TechCrunch also paraphrases Ronacher on model routing: using an LLM to pick a model can be too expensive; Jev’s cost/speed make real-time sorting more plausible. That is the reporter’s summary.

Founder voice (not a customer testimonial)

Diogo Almeida, TypeSafe founder/CEO, in the same TechCrunch piece:

“We have lightning in a bottle, and yet it is not useful… The problem is we are optimizing for human language … We have been super good at human language for four years, but it’s not useful for automation because computers speak a different language.”

On synthetic data / RLCD (same article):

“We made an early bet that we will be making all of our data, and that has been one of the best bets I’ve ever made in my life — better than our launch, in my opinion, better than RLHF.”

Launch-post byline (15 September 2026) is Almeida’s product narrative, not a third-party review — founders.

Investor voice (BusinessWire, 15 September 2026)

James Hardiman, General Partner at DCVC, in TypeSafe’s seed announcement — investor, not a product user:

“TypeSafe is approaching one of the biggest remaining challenges in AI: turning increasingly capable models into technology that developers can reliably build into products at scale.”

Funding facts belong on TypeSafe AI. A seed quote is not a customer case study.

What we will not put here

If you have a public, named, dated write-up, it belongs on the reviews roundup as an outbound link — not rewritten as our testimonial.

FAQ

Are these jev.pro endorsements?

No. They are excerpts from public articles so you can click through.

Why so few quotes?

Because we only repeat named public speech. Thin pages of fake praise would fail the field-guide bar.

Can I submit a testimonial to jev.pro?

This is not a testimonials CMS. Publish somewhere citable; we may link it from the roundup if it is public and specific.

What this page does not claim

Sources