The framework, the rubrics and the human review, in the open.
What Competency Labs measures, how a verified result is produced, and what it does not tell you.
What exists today
- Practice Labs in the public catalog
- 13
- Roles with at least one Practice Lab
- 12
- Published role guides
- 14
Counted from the catalog on this site. One role, Clinical Research Coordinator, has an open verified pathway today.
The method
- People do realistic work in a fictional study, Glycoterra: they read protocols, consent forms, source and query records, then decide and write.
- We capture what is observable: decisions, action order, work products and communication, including SBAR where it fits.
- Practice is low stakes and formative. Its scores never become verified evidence.
The competency framework
- The published standard, BORAKA-VCS-CR-001, defines ten competency domains and twenty canonical competencies across twelve registered role families.
- Each competency is described at four behavioral performance levels, so a result says what was observed, not just that a topic was covered.
- Five Evidence Assurance Levels describe the strength of a competence claim, from practice work up to human-reviewed verified assessment.
- Six evidence types are recognized: knowledge, decisions, action sequences, artifact review, work products and communication, plus reviewer observation.
- Results distinguish not assessed, developing, demonstrated and verified. Not assessed is not a failure; it simply means we have not seen that behavior yet.
Read the source in the standard
Rubrics and versioning
- Every verified attempt is scored against a named rubric version. The current version is BRK-RUBRIC-2026.1; the standard it applies is BORAKA-VCS-CR-001 v1.0.0.
- A credential always carries the version it was reviewed under. When the rubric changes, older records keep their original version so a result can always be reproduced and audited.
- Each task is tagged to a role, a competency, the behavior we expect, the evidence type, a rubric and a difficulty before it is used.
- Detailed scoring keys, planted issues and calibration materials are not published, so an active assessment cannot be gamed. The structure of the rubric is public; the answer key is not.
Read the source in the standard
Human review
- Practice work is never sent to a reviewer. It is developmental only and produces no employer-facing record.
- A verified assessment is reviewed by a trained human assessor against the published rubric version. AI may assist with drafting; it never issues a credential.
- A quarter of finalized verified reviews are read a second time for calibration, so assessors stay aligned. The selection is fixed by the system and cannot be rerolled by a reviewer.
- Reviewer identity is masked on the candidate-facing record as an opaque reference. Reviewer role, rubric version, timestamps, scope and limitations are recorded.
- Accommodations are available, and a candidate can ask for a result to be reconsidered.
Read the source in the standard
Validity and renewal
- A verified result carries a two-year validity window, then reassessment. Reminders go out before the window closes.
- Earlier results stay on the record as earlier snapshots. A renewed result never overwrites the history of how someone improved.
- An expired result is shown as expired. It is never deleted and never presented as current.
Read the source in the standard
In progress, not yet published
Two pieces of work are underway. They are listed here so you know what is coming; neither exists yet, and we will not reference them as evidence until they do.
Coming soon
Role benchmarks
Aggregate, anonymized score distributions per role, published only once the underlying assessment volume is large enough to be meaningful.
No figures yet: there are not enough real, human-reviewed results. The methods paper sets out the rules before the first figure exists.
Download the methods paper (PDF)Email me when the first benchmark is published
Coming soon
Framework crosswalk
A public mapping of our competency framework to the published frameworks we reference, so employers can see exactly where each behavior comes from.
Independence and limits
BORAKA-VCS-CR-001 is aligned at a high level with ICH E6(R3) Good Clinical Practice and is designed for crosswalk to the Joint Task Force for Clinical Trial Competency Framework. Referencing a framework does not mean the body behind it endorses, accredits or partners with Competency Labs. Unless a written agreement says otherwise, no regulatory agency, professional association or standards body has reviewed or approved this standard.
The standard is published for public and subject-matter-expert review. It is not an accreditation, a professional license, a regulatory approval, a certification by an independent standards body, or a claim of universal employer recognition.
A result describes the assessed work only: it is not a regulatory qualification and does not predict job performance. Employers read it alongside experience, qualifications, interviews and supervision needs, and they remain the decision-makers.
Read the source in the standard