Blog · Published · Updated
How to choose AI for translation in a company?
Short answer
You choose AI for translation on the material your team actually prepares, counting the work needed until the translation is accepted. The assessment covers meaning, completeness and company terminology, because fluent text can drop a delivery condition or change a unit of measure. You check the glossary on a specific term.
- net for implementation for a team of up to 10 people, paid once
- from €1,150
- from start to work of the entire team
- 2 weeks
- after this time, we measure the effect on the same tasks
- 30 days
- conversation with an engineer and a written plan within 2 business days
- €0
We belong to the OpenAI and Anthropic partner programmes and learn through their official courses. What this means ↓
Author: Syntalith · 4 min read
Statistics
Translation and AI: statistics 2026
0.7 mo.
payback of a €1,150 training day for 8 people saving 1.5 hours a week each: €1,813 of time recovered against €143 of licences a month. Scenario.
40%
According to a study published in Science (2023), ChatGPT cut the time 453 professionals spent on writing tasks by 40%.
Source: Noy and Zhang, Science (2023) ↗
48.8%
According to Eurostat (2025), 48.8% of EU companies that considered AI but did not use it named concerns about data protection and privacy.
20.0%
According to Eurostat (2025), 20.0% of EU companies with at least 10 employees used AI.
Source: Eurostat, 20% of EU enterprises use AI technologies (2025) ↗
48%
According to the European Commission (2026), 48% of small and medium-sized companies in the EU name a lack of expertise as a barrier to AI.
Source: European Commission, Flash Eurobarometer 578 (2026) ↗
Figures checked: 22 October 2026
- Texts the team really translates
- A message with a delivery condition and units
- Company terms in the glossary
- The time to an accepted version
Fit the sample to the team's work
For an export department, a product description, a message with delivery terms and a reply to a client's question are useful. In technical documentation, recurring part names and references to tables may matter more. Choose material cleared for use in the tools you test. Fictional texts can be prepared for the first trial.
Name a person who knows the target language and the subject of the document. A verdict that the translation „sounds good” is not enough. The reviewer has to be able to check whether a condition or an instruction kept its meaning. If you have no such person, count the cost of their work before you run the comparison.
Record the text version, the tool used and the date of the trial. Also keep the instruction and the glossary if you add them. Comparing two results prepared with different materials will not show where the difference came from.
Check the glossary on a specific term
A glossary is an agreed list of terms and their equivalents. It can cover part names, job titles or elements of a service. Also list proper names and codes that must stay unchanged.
DeepL describes applying a glossary to file translation, provided the source and target languages match the glossary. Check whether this feature is available in your environment. We read the instructions on 9 October 2026. Applying a glossary to file translation in DeepL.
Having the feature does not settle the comparison. Put the chosen term into several sentences and check whether its use fits the context. The list of accepted equivalents may also be out of date. Someone on the company's side should approve it before the test.
A fictional message with a delivery condition
Suppose the source text describes 24 packs of 6 units each. Shipment will take place within 5 working days of payment being confirmed. The message gives no date for the goods to arrive. This is made-up material for the exercise and describes no real order.
The reviewer should check whether the translation keeps the number of packs and units, tells shipment from delivery and keeps the condition of payment confirmation. A promise of delivery within 5 days would change the content. So would turning 24 packs into 24 units.
| Area | Question for the reviewer |
|---|---|
| Meaning | Did the translation keep the commitment and the condition for meeting it? |
| Completeness | Does every important piece of information in the source have an equivalent? |
| Figures and units | Are the quantities, dates, currencies and units still correct? |
| Terminology | Were the accepted terms used in the right context? |
| Usability of the file | Do the headings, tables and labels allow further work? |
Mark an error that changes a commitment separately from a style correction. Do not add them up into one number without an agreed weight. A tool with few small corrections may be unsuitable for a given process if it repeats an error in order terms.
Count the time to an accepted version
Measure preparing the material, getting the translation and correcting it. To the cost of the tool, add the work of the person who approves it and the upkeep of the glossary. The speed of generating text describes only part of the process.
Repeat the assessment on new material. That shows whether an improved instruction helps beyond the first example. If both tools need similar correction, the choice may come down to how convenient it is to work with files, the access terms and the company's data rules.
When to order training or an integration
When the team needs to agree how to prepare material and check translations, Syntalith team training can be the starting point. The process review at Syntalith is free, automations, apps and agents start from €3,000 net, and team training from €1,150 net a day (pricing). We set the scope on the company's tasks. Training does not replace the linguistic acceptance of documents, and it does not confirm the quality of the chosen tool for every use.
If approved translations have to be moved regularly into several systems, assess process automation separately. A quote will need the sources, formats, places where files are saved and the approval method. A glossary or a larger number of texts alone does not settle the need to build an agent.
What it costs
- Time saved
- 8 people × 1.5 h a week × 4.33 = 52.0 h
- Value of that time
- €1,813
- Licences
- $160 (€143)
- Left after licences
- €1,670 a month
Payback of a €1,150 training day
0.7 months
The hours saved are the scenario's assumption, which we check by timing tasks before the training and after 30 days. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat. Claude Team seats at $20 per person a month, at the ECB rate of 1.1205 USD per euro on 22 October 2026.
- The time of the task before the training and after 30 days
- The corrections made by the person checking the result
- Tasks where a participant still needs help
- Licences and tools cleared for work
Where to start
We run the workshop as part of team training, from €1,150 net a day. For the free review, a description of one task the team does every week and the tools it uses is enough. You receive the written training programme within 2 working days.
„I already use AI.”
Your people too. Just no rules.
From our conversations with companies
5 na 10 employees paste sensitive data into private AI accounts: contracts, invoices, customer data.
In ESET survey (2026)
1 na 10 admits this in the survey.
Customer data pasted into a private AI account without an data processing agreement is a violation of the GDPR.
Don't worry, we won't tell anyone :)
What do we see in companies that "already use"
- Only ChatGPT or Claude accounts, without plugins, skills, integrations and MCP servers with company tools.
- Company programs are not connected to AI. People still click through systems and Excel, and enter chat data manually.
- Lack of company standards: everyone writes instructions in their own way, no one shares what works.
- No rules for personal data, GDPR and company secrets.
- No limits and no visibility: no one knows who uses AI, how much and for what.
62% Polish employees use AI at work · 27% companies have written rules for working with AI.
Source: ESET and DAGMA, Cyberportrait of Polish business 2026 (March 2026, 1,026 people) ↗
ChatGPT or Claude space review
Accounts, permissions, rules for working with data and 5 tasks set for the team. 1 day.
€450 net
Company accounts, rules of working with data and team training in 2 weeks, from €1,150 net.View implementation for the team →
„Is my data safe?”
Yes. You choose where they go: from your company account to the model on your server.
00Employee's private account
Customer and company data ends up in an account that the company cannot see or control. There is no data processing agreement, so it is a violation of the GDPR. When an employee leaves, the chat history goes with them.
01ChatGPT or Claude corporate account
The provider does not use your data to train models. A copy of the queries remains in his logs for up to 30 days.
02API with zero retention (ZDR)
The provider processes the request and does not save anything. It requires his consent, we prepare the application together.
03Processing in the EU
Queries are counted and stored in Europe: OpenAI in the EU region, Claude via clouds with the EU region.
04Model at your place
The open model runs on your server or in your cloud in the EU. The data does not go outside the company.
What we do with every implementation
- Confidentiality agreement and data processing agreement before we receive any data.
- You choose the cases for the prototype and we anonymize them upon request.
- The data stays in the EU and we delete it after the test.
- Corporate accounts with an administrator on your side.
When data cannot leave the company
The open model runs on your server or in your cloud in the EU. To choose from:
- White-tailed eagle
- Polish model, SpeakLeash and ACK Cyfronet AGH
- PLLUM
- Polish model, HIVE consortium commissioned by the Ministry of Digitization
- Qwen
- Alibaba
- GLM
- Zhipu AI
- DeepSeek
- DeepSeek
We select a model for the task and check it on your cases. We run it, tune it and maintain it.
Source: Bielik and PLLuM, ACK Cyfronet AGH ↗ · PLLuM, Ministry of Digitization ↗
Do you have a lot of your own data? We will fine-tune the small model.
- 1 Your examples
- 2 tuning
- 3 test on new cases
- 4 model at your place
When you have thousands of similar emails, documents or reports, we teach a small open model based on your examples. It works for you, responds quickly and costs less to maintain than a large model. We measure quality on issues that the model has not seen while learning.
Przemysław and Piotr did it in large programming and consulting companies. Artem did it before and does it today at Syntalith. It's about models on your own infrastructure, tuning and quality measurement.
You choose the protection level during the interview. We sign the data processing agreement before we see any data.Schedule a free analysis →
Price
Three packages. You know the price right away.
Team of up to 10 people
from €1,150
net, paid once · 2 weeks
Company, from 11 to 40 people
from €2,300
net, paid once · 3–4 weeks
Organization, over 40 people
from €4,600
net, paid once · 6–8 weeks
- The same packages once you have licenses. To start, an inspection of the space in 1 day is enough: €450 net.
- You buy Claude Team licenses directly from Anthropic, without our markup: USD 20 per person per month for annual payment or USD 25 for monthly payment, from 2 to 150 people.
From conversation to measurement
01 · 30 min
Free process review€0
An engineer picks 3 to 5 of your team's tasks with you. Written plan within 2 working days.
02 · 2 weeks
Rolloutfrom €1,150
Accounts and roles, rules for working with data, training on your team's tasks. You buy licences directly from the provider.
03 · after 30 days
Measurementincluded
We compare task times before and after. We name in writing the tasks worth handing to a system.
or call+48 888 78 48 78WhatsAppMon–Fri 9:00–17:00 CET
Worked example · Scenario
How we count the payback: translation in a company
We run the same count for your team in the free review.
€3,490
a month for quoting 150 enquiries by hand
Payback: 1 mo.
40 minutes to quote a request and checking in 15% of the old time are our assumptions. €34.90 an hour is the EU average hourly labour cost for 2025 according to Eurostat, for companies with 10 or more employees. Model cost: reading the enquiry in our quoting system (about €0.0008) plus pricing the lines (Claude Sonnet 5.5, about 4,000 + 1,500 tokens), at the ECB rate of 1.1205 USD per euro on 22 October 2026. The extra margin is a scenario: we assume 1 in 40 enquiries quoted the same day ends in an order that a slow reply loses today, with a margin of €344. In the review we count it on your sales process.
See the full calculation →Questions
How do you choose AI for translation in a company?
On a sample of the team's material, assessing meaning, completeness and terminology and the time to an accepted version. You check the glossary on a specific company term.
Is a glossary in a translation tool enough?
It keeps terms consistent, but it does not check whether the translation kept the conditions and the figures. A person who knows the content judges that.
How much does this training cost?
At Syntalith, from €1,150 net for a workshop day for the team. Claude Team seats cost $20 per person a month. With 8 people and 1.5 hours a week each, the payback is 0.7 months.
