From practice.Built for practice.

From a pattern I met from both sides. In the police service I ran the search for an interpreter myself; as a sworn interpreter I was at the other end of it. The cause was the same every time: not a lack of willingness, but a lack of availability. At some point it was clear that no further list would change that, only an infrastructure.

Over fifteen years of experience in interpreting. After passing the military language mediator examination at the Federal Office of Languages in 2011, I served as a language mediator and linguist in a military context, at non-commissioned officer rank, in Germany and abroad. I then worked as a police officer with the French Police Nationale.

After leaving public service I gained two Certified Translator qualifications, the Master Professional in Business Administration awarded by the Chamber of Commerce, and a Bachelor of Arts in Business Administration.

Since 2017 I have been a generally sworn translator for German, French and English at the Regional Court of Saarbrücken. I am also registered as a translator and interpreter for German and French at the Tribunal Judiciaire de Sarreguemines.

Over 2,000 hours in law enforcement and the courts, from minor offences to capital crimes. Entry, asylum and deportation hearings. Custody hearings through to main trials. House searches through to the execution of arrest warrants. Telecommunications surveillance. In almost every one of those hours a decision was at stake that had consequences for somebody.

A station enters the language, the estimated duration and the case number and sends the request. It goes out in stages to those who meet the language and minimum qualification and are available. Whoever accepts first gets the assignment. Both sides enter an end-to-end encrypted video room, having first confirmed the notice on Section 201 of the German Criminal Code.

The authority sets the rules itself: its own pool or nationwide, minimum qualification from language mediator through to sworn, allocation from random through rotation to direct assignment. If the connection drops, both sides can re-enter the room. After the assignment a PDF record is produced with duration, language and case number, which serves as the billing basis and fits into the file.

No, and that is not a statement of intent but a property of how it is built. Payment follows the JVEG where it applies, otherwise the rate agreed with the authority. The rate appears in the request before anyone accepts. There is no bidding process, no undercutting and no ranking that could be improved by lowering prices.

The contractual relationship is between the authority and the interpreter. No fee income flows through us. We earn a service fee, collected separately and only once an assignment has been completed.

For people starting out. Access to public-sector interpreting runs on established patterns: lists, personal contacts, recommendations. Newcomers barely get in, and not because they lack qualifications, but because the settled relationships between authorities and interpreters leave little room. Random allocation changes exactly that: it asks about language, qualification level and availability, not about how well known you are. Anyone who meets the requirements gets the same chance of a request as someone who has been on the list for twenty years.

Because it does not have to trust us; it can check. Before any decision we provide the documents: architecture overview, methods used, data centre details and the draft data processing agreement. Where a requirement is not met, we say so instead of writing around it.

Then there is the choice of operating model. Anyone who does not want us inside their own infrastructure gets a dedicated instance or full operation within their authority network. The authority decides the model, not us.

Federal and state governments are building their own AI models for public administration, and translation ranks high on the list. These systems are coming, regardless of what the profession thinks of them. What decides whether they are used in interviews is rarely quality. It is availability.

If nobody can be reached at three in the morning, the machine wins the argument without the argument ever being had. That is why lingomatch makes qualified people as available as a machine would be. Every assignment we place is evidence that the alternative works.

Because interpreting in a public-authority context is a profession built on trust. The authority trusts that the rendering is accurate. The accused trusts that they are being understood. The interpreter trusts that they will be treated and paid fairly. If a platform does not take that trust seriously, it has no right to exist.