Specialist vetting should produce a current capability record that can be matched to a real assignment. Evidence of domain skill, language, security training and availability needs an expiry date and a clear review route before a worker is marked ready.
Composite reference architecture. A modelled operating system; no hiring or placement metric is represented as realised.
The operating problem
A résumé does not prove task readiness. Delivery teams need current evidence of domain skill, language, security training, availability and performance on the exact workflow.
How the workflow works
Match the project requirement to task-specific assessments and verified credentials. Keep availability separate from competence: a strong specialist can still be unavailable or overbooked. The capability record should show when each check was completed, what it covered and whether retraining or reassessment is due.
Where review and controls matter
A talent lead reviews identity concerns, accommodations and high-risk assignments. Weak proxies should not determine readiness. When early production quality disagrees with the assessment, review the assessment design as well as the worker's performance. Placement, retraining and holds need an explicit reason and an appeal path, so an old assessment does not silently exclude someone from future work.
Risk reference: NIST AI RMF.
What to measure
Use a fixed cohort and measurement window. Count numerator events only within the denominator's eligible cohort; sampled measures use the same audit sample for both terms. Record the source, exclusions and owner, and keep missing data visible. These definitions are proposed measurement contracts, not observed results.
| Measure | Definition and unit |
|---|---|
| Time to verified capacity | Time from approved staffing need to the required ready specialists becoming available; median and p90 days. |
| Match acceptance | Proposed assignments accepted by both delivery owner and specialist divided by eligible proposed assignments; percent. |
| Early quality | Audited tasks passing the rubric during a fixed early-assignment window divided by audited tasks in that window; percent. |
| Redeployment | Eligible released specialists placed in another suitable assignment within a fixed follow-up window divided by eligible released specialists whose full window has elapsed; percent. |
| Appeal resolution | Appeals resolved within the promised deadline divided by appeals due; percent. Report time and outcome separately. |
A useful first pilot
Match a small project demand to an existing specialist pool, including expired credentials and overlapping availability. Review early task quality against the exact evidence used for placement.
Bring the workflow, its source records and the decision that needs an owner to a BergLabs pilot discussion.
