Definition
Concentration of complex test execution across qualified analysts, computed as a normalized sum of workload shares. Rises as AI scheduling routes hard work to the fastest analysts and the rest of the qualified bench decays.
The Measurement Gap
AI lab scheduling routes complex tests to the fastest, most reliable analysts because the algorithm optimizes for throughput and minimal repeat analyses. Over time, two of six qualified analysts handle 68% of complex work. Utilization metrics look balanced (everyone is busy) but the complexity distribution is severely skewed. The remaining four qualified analysts lose proficiency from disuse, widening the gap. When a key analyst takes vacation or leaves the organization, complex test turnaround time spikes because the backup analysts haven't maintained the skills. No current metric tracks what kind of work each analyst performs, only how much. The skill concentration risk (the operational consequence when key people are unavailable) is completely invisible to standard lab performance metrics. The AI creates a self-reinforcing loop that the organization cannot see and that traditional workload metrics actively obscure.
Decision Supported:
If the top two complex-test analysts are unavailable tomorrow, how much of the lab's complex testing capability disappears, and is AI routing making that concentration worse over time?
Formula
Notation:
ASCI = Σ (s_j)² for each analyst j
ASCI_normalized = (ASCI − 1/N) / (1 − 1/N)
Variables:
- s_j = share of complex test executions performed by analyst j
- N = total number of analysts qualified for the complex test type
Worked Example: Six analysts qualified for advanced dissolution testing. Last quarter: Analyst A 34%, B 32%, C 14%, D 11%, E 6%, F 3%. ASCI = 0.34² + 0.32² + 0.14² + 0.11² + 0.06² + 0.03² = 0.2542. ASCI_normalized = (0.2542 − 1/6) / (1 − 1/6) = 0.105. Below the 0.30 threshold but the top-2 share of 66% is concerning.
Edge Cases:
- ASCI_normalized = 0 means perfectly even distribution across all qualified analysts
- ASCI_normalized approaching 1.0 means one analyst monopoly
- Track the "top-2 share" companion metric alongside ASCI for practical key-person risk assessment
- Calculate per complex test type: concentration may vary by testing discipline
Sign convention: Lower ASCI_normalized is better.
Proposed formula, pending validation pilots.