AI evaluation and data
- Rating and correcting model responses
- Prompt and test-case writing
- Annotation and language QA
- Red-teaming in the target language
Language bar: C1 or JLPT N2, plus a sample evaluation task
For employers
Tested Spanish, French, German and Japanese speakers who join your team remotely, use your tools and work your hours. Billed by the hour, starting with a pilot of 2–5 people.
Language bar: C1 or JLPT N2, plus a sample evaluation task
Language bar: B2 or JLPT N3
Language bar: B2 or JLPT N3, plus domain training
Language, level, tools, working hours and how many people you need.
Within about five working days, with verified language levels and sample work.
Talk to the people you like and pick your team.
They work your hours from India. We handle payroll, equipment and replacements.
People who already have the language and the domain skill. Best when you need someone in one to two weeks.
Strong language graduates who complete a four to eight week bootcamp in your kind of work before they start.
Your team, our payroll. We manage contracts, equipment and continuity; you manage the work.
Start with a pilot of two to five people on one project. If someone isn't the right fit, we replace them.
Plan a pilotBy the hour, per person. The rate depends on the language, the level and the kind of work. AI evaluation roles are priced higher than general operations roles. We share rates once we understand the role.
From home in India, during your working hours: US, European or Japanese time zones.
Fluentras does. People are on our payroll and contracts; you manage the day-to-day work like any member of your team.
Tell us and we replace them. Pilots are designed to be low risk.
Anonymised profiles within about five working days of a clear brief, and people at work in about two weeks.