Train AI.
Evaluation and annotation in a professional field, done by assessed people and delivered with a record you can check yourself.
Why expert judgement in Albanian, from betaPlan
- A low-resource language
- Albanian has little training data and few annotators who know both the language and a professional field. Someone who only knows the language will not correct a model that gets accounting or pharmacy wrong in Albanian.
- Professionals, not a crowd
- The work is done by accountants, pharmacists, lawyers and other professionals qualified in their own field, not by casual workers from platforms where anyone can pick up the work.
- Two assessors, a published rubric
- Two independent assessors decide whether each annotator qualifies, on a work sample and against a rubric published with a version number. No decision is taken by AI and none comes from self-assessment.
- A record for every piece of work
- Every task comes with a record of where it came from: who did it, who reviewed it, under which guideline and with what decision, sealed so you can check it yourself without asking us.
The record itself
Batch BP-2026-0417Field HealthcareTask T-0910
Discharge note reviewed. Metformin dose stated as 1000 mg twice daily is correct for this patient weight and renal function. No interaction flagged with the listed antihypertensive.
- Licence category
- Pharmacist
- Reviewer identifier
- bench/AL-3F92
- Independent review
- bench/AL-7C41
- Unaided
- Yes
- Review outcome
- approve
- Sealed at
- 2026-04-17T09:14:02Z
- Record hash
- 5d9c43d3da…aa93fd
This delivery seals 3 records. The outcomes are approve, reject, approve with correction. Each one was reviewed by a reviewer whose identifier differs from that of the person who annotated it. The record that failed review is sealed into the delivery like the others, in the same chain, rather than removed from it.
The field, licence and note above appear exactly as sealed in the record, unchanged and in the language they were written in; the record's hash covers those exact words.
In this section
AI companies
Evaluation by professional field, documented as evidence: the field, the assessed capacity, the work we run, the independent review and a record a buyer can check.Human evaluation
How learning leads to qualification: the decision path, two named assessors for every work sample, and a visible assessment rubric with a version for each programme.Verified records
What a delivery contains, how to check a record offline with a published key, and what a failure looks like.Case studies
Real cohorts in Albania from 2021 onwards, with the figures of each engagement.
Where to start
With the kinds of work we do, or with how our assessors qualify.