Astana · Almaty · Tashkent · Belgrade
Verified AI trainers from Central Asia, the Caucasus and the Balkans
Native Kazakh, Uzbek, Kyrgyz, Serbian and Russian speakers for RLHF, evaluation and annotation. Each one is ID-checked, sanctions-screened and interviewed live by a named person. They work inside your tools, and we pay them locally within 48 hours.
On every profile
You get the evidence, not just a badge
Each worker’s profile shows who checked what, and when. Document numbers stay hidden; the result and the date are always there.
- Identity: provider, date and document type
- Residency and sanctions: country, last screening, result
- Languages: test score, proctored or not, interviewer, date
- Track record: hours billed, QA score trend, client ratings
Evaluator · ID 2417
Kazakh · Russian · Astana
- Identity
- Sumsub · national ID card · 12 Oct 2026 Document number never shown
- Residency and sanctions
- Kazakhstan · screened 28 Oct 2026 · clear OFAC SDN, EU, UK lists
- Languages
- Kazakh 91 · Russian 94 · both proctored Interviewer A. S. · 14 Oct 2026
- Skills
- RLHF ranking 86 · annotation 90 Label Studio, CVAT
- Track record
- Probation: 6 of 10 hours reviewed QA score shown after 30 days
Vetting
Five gates before anyone bills an hour
The first three are automated and free for the candidate. The last two are checked by a person. After that, every new project starts with a 10-hour probation.
-
Sign-up screen
Residency self-certification, IP check, sanctions name screen, device and phone check.
-
Native-language test
Written in the language, timed, with register and dialect items and paste detection.
-
Skill test
Annotation, RLHF ranking or a domain quiz, scored automatically.
Tier 1 · never sold -
ID check and proctored retest
Document and liveness check, then a recorded retest on new items.
Checked by a person -
Live interview
15 minutes with a native interviewer, including a 5-minute work sample on a shared screen.
Checked by a person Tier 2 · ready to hire
Probation: the first 10 billed hours on every new project are double-reviewed.
Tier 3: 200+ billed managed hours with a QA score of 92% or more.
Rates
Published rates, billed in USD
Managed teams are billed by the hour from this card, from $3,000 a month per team. Calibration, probation reviews and the 10% QA sample are billed at the same rate and itemised on your invoice.
All pricing| Role | Managed rate, USD |
|---|---|
| Kazakh, Uzbek, Kyrgyz native QA and alignment | $24–30 /h |
| Serbian / BCS native evaluator | $18–22 /h |
| Russian annotation and evaluation | $12–14 /h |
| Domain expert: engineering, medicine, law, finance | $40–55 /h |
| Software engineer: code evaluation, RLHF | $32–45 /h |
Compliance
Answers for your legal team before they ask
- Sanctions
- Screened against OFAC SDN, EU and UK lists at sign-up and every month.
- Residency
- Proof of address and an IP check at every login. A VPN login from a sanctioned country suspends the account.
- Personal data
- Stored in Kazakhstan, as Kazakh law requires. Identity-document images are never stored.
- Your data
- Workers operate inside your tools, so your data never enters SilkTrain systems.
Pilot
Start with a paid two-week pilot
10 vetted workers in your tool, ready within 10 business days (5 for Russian). If a managed team’s weekly gold-set agreement falls below 90%, the hours below it aren’t billed.
For trainers
Speak Kazakh, Uzbek, Kyrgyz, Serbian or Russian?
Paid AI-training work from home. Free tests that take under 45 minutes, and payouts in your local currency within 48 hours of your weekly timesheet being approved.