AI Translation Workflow Consulting
Audit, Build and Run: from the first decision to day-to-day operation
|
▮Where are you now?
- You are considering AI translation but have not decided what it should handle. Start with Audit.
- You have decided, but your tools, translation memories and content systems are not ready. Build.
- You publish content every month and do not want to manage the workflow yourself. Run.
Only need one AI translation checked before it is published? That is Independent Review, not consulting.
▮Three packages. Most clients start with Audit.
Each package can be bought on its own. Together they follow the order in which this work usually happens: decide, build, operate.
| Audit | Build | Run | |
| The question it answers | What can go to AI, and what must stay human? | How do we make it work in our own tools? | Who keeps it running and checks the quality? |
| What you get | A written report: your content sorted by risk, where the checks sit, a cost comparison and a roadmap | A tested workflow in your CAT tool or platform, connected to your content systems, documented and handed over | AI draft and human review in every language, terminology kept current, a monthly quality report |
| Duration | Two to four weeks | Agreed in the written scope | Ongoing |
| Pricing | Fixed fee | Fixed project fee | Monthly retainer or per release, no per-word invoicing |
▮AUDIT: Decide what goes to AI and what stays human
We look at what you translate, where an error would cost you, and how your current output is checked. You receive a written workflow your team can follow and your clients can audit. If your in-house AI translation already works and you can show how it is checked, we will tell you there is nothing to fix.
▮BUILD: Set up the workflow in your own tools
Once the decisions are made, the workflow has to work with your tools, your translators and your deadlines. We set it up in your environment: AI drafting inside your CAT tool with your translation memory and glossary, automatic checks before a reviewer sees the text, and connections to your content systems where they are feasible. You receive a tested workflow, its documentation and a briefing for your team.
▮RUN: We operate the workflow for you
For companies that publish content continuously and do not want to build an in-house language team. We run the AI draft and human review in every language, with a named reviewer per language, keep terminology current as your products change, and send you a quality report every month. One point of contact for every language.
▮What we do not send to AI
Medical device documentation, texts with legal effect and safety-critical instructions run through our fully human workflow (HumanExpert) by default. Where a sworn translator's signature is required, the document stays with a sworn translator. An audit that recommends AI for everything is not an audit.
How each stage works in practice, from content classification to API connections and automated QA: Building an AI translation workflow: what happens at Audit, Build and Run
|
|