Skip to content
About

Trust, manufactured honestly.

Prevouched exists because the offshore software market has a pricing problem that has nothing to do with price. Buyers are paying a quiet risk premium on any team they cannot put a name and a phone number to. Good shops lose deals to mediocre ones who happen to share a time zone with the buyer. That is a market failure with a knowable shape. And it is the only one we work on.

We are not a marketplace, not a staffing firm, not a delivery shop. We audit offshore software agencies against a public rubric, list the ones that pass, and put a US-based contact on the calls where the relationship is most fragile. So a credentialed offshore team can be hired the way a credentialed onshore team already is.

Operating principles

Six rules we will be held to.

P-01

Asymmetry is a product problem, not a discrimination problem.

Buyers don't pay a risk premium on offshore teams because they want to. They pay it because they have no instrument to discharge the personal-reputation risk of being wrong. Build the instrument and the premium collapses.

P-02

Trust that cannot be lost is not trust.

Every claim Prevouched makes is bound to evidence on file and a revocation path that is published. A badge that doesn't track liveness is a logo; a rubric without consequences is marketing.

P-03

The money flow is the first ethical question.

We only ever invoice the agency. We are never a payee on the agency↔client contract. The day we take delivery margin is the day our endorsement is for sale.

P-04

Publish the rubric. Keep the keys.

The categories, weights, pass bars, and revocation triggers are public. The internal scoring keys are not. Because the moment scoring keys are public, applicants train to them and the rubric loses signal.

P-05

Accountability is a product, not a vibe.

A named US liaison, a written SLA, and a covered-call definition do more to close trust gaps than a hundred case studies. We sell the named person; we don't pretend they replace the buyer's procurement.

P-06

Refuse the easy sale.

We turn down agencies that would pay full freight if accepting them would dilute the mark. The price of an endorsement that cannot be bought is foregone revenue. That's the trade.

What we will not do

The shape of the business is in the refusals.

Most of these are obvious revenue we have already turned down. Saying them in public is how we tie the hands of our future selves.

Sell delivery.
Sharing economic interest in delivery would invalidate the endorsement on every engagement we touched.
Take a markup on the agency↔client contract.
It would put us on the wrong side of the buyer's interests during a dispute. Membership + bounded success fees only.
Issue irrevocable marks.
The whole product is the revocability. An irrevocable badge is a logo.
Run a pay-to-list directory.
Placement weight is earned. The directory's value is the credentialing behind it.
Promise outcomes we cannot guarantee.
The mark is admissions, not insurance. Outcomes belong to the agency and the buyer.
Use the badge as a marketing widget.
It's a credentials document. We resist any design choice that would let it read as a press graphic.
How we got here

Short history. Documented intentionally.

2025 · Q4
Initial intake interviews with 60+ offshore studios across SE Asia, LatAm, and EE. Loss patterns converge to four. The same four every shop names.
2026 · Q1
Rubric v0.1 drafted and stress-tested against ten anonymized historic engagements. Pass bars calibrated against a reference panel.
2026 · Q2
First cohort vetted. Liaison model piloted on three Backed engagements; two reach milestone gate without escalation.
2026 · Q3
Public rubric v1.0, public revocation policy, public changelog. Directory opens with the first batch of marks.

If any of this rings false, tell us. The shortest way to improve the rubric is to be shown the thing it misses.