Own the translation
Running our own models is harder and more expensive than calling an API. It is also the only way to promise that your content never reaches a third-party AI service and have that survive a security review.
About
Built for documents that can't leave
SpectrHQ is private translation infrastructure for organisations that handle confidential content - legal, medical, financial, technical, and anything else that cannot be pasted into a public AI tool.
Why it exists
The problem was specific
SpectrHQ started because an organisation needed to translate confidential documents and every available option meant handing them to a third party. Machine translation services process content on infrastructure the customer has no visibility into. Agencies distribute files to freelancers. Workflow platforms route text through whichever engine they have integrated that quarter. For content that cannot leave the organisation's control, none of those were acceptable.
So the answer was to build the translation itself - models operated directly, on infrastructure under our own control, with no outside AI service anywhere in the chain. Everything else in the product exists because a document is not finished when it comes back translated. It still has to be reviewed, corrected, approved, and shipped, and that is where organisations actually lose their time.
How we build
Four decisions that shaped the product
Meter words, not people
Charging per seat means rationing access to the reviewers and subject-matter experts who make translation correct. Editors, reviewers, and projects are unlimited on every plan. We count translated words and nothing else.
Two plans, not five
We publish two self-serve plans and one custom agreement. When a tier exists only to make the tier above it look better, it is a pricing tactic rather than a product.
Publish controls, not configurations
We state what we guarantee and we do not publish our vendors or our internal architecture, because that detail helps an attacker more than it helps a buyer. Under NDA, we go as deep as your security team needs.
Who builds it
No support tier between you and the product
SpectrHQ was founded by Matteo Galiano, who leads its development and handles enterprise scoping and security review directly. Send a questionnaire and it is answered by someone who knows how the system works, not routed through a queue. Requirements that matter to your organisation reach the people deciding what gets built next, without passing through three layers first.
Matteo Galiano on LinkedInWhere we are today
Early, and specific about it
SpectrHQ is a young company. Our SOC 2 programme is under way with Type II expected in 2027, and we run on SOC 2 Type II certified infrastructure today. We would rather tell you exactly where the programme stands than imply more than we have - the same reason our security page describes what we guarantee instead of what we intend.
What that buys you now is access. Enterprise agreements are scoped directly with the people building the product, security questions are answered rather than deflected, and requirements that matter to your organisation can still shape what gets built next.
Company
The details
- Legal entity
- SpectrHQ is a product of MVG Projects LLC, a US-based limited liability company.
- Founded by
- Matteo Galiano
- Contact
- hello@spectrhq.com
Get started
Put a document through
Start free and translate today, or bring us your workflow and security requirements and we will scope the workspace around them.
14 days. 5,000 words. No card required.