In the first year under the structured AI training method, SES/CONT's average case processing time fell 18.2% (17.92 to 14.66 days) while documentary output per case stayed stable¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? SES/CONT's average processing time fell from 17.92 days in 2023 to 14.66 days in 2024, an 18.2% aggregate reduction, with documentary output per case stable (2.77 to 2.78). → Vinicius Santana Gomes 2026

Evidence¶

Vinicius Santana Gomes 2026¶

Vinicius Santana Gomes. (2026). The Main Barrier to AI Adoption in the Public Sector Is Lack of Training: How a Structured Method Accompanied Productivity Gains in Two Brazilian Government Cases. arXiv. https://arxiv.org/abs/2606.01517 doi:10.48550/arxiv.2606.01517

q2 · i? · design · r2

Observational before-and-after comparison of official SEI-GDF statistics for SES/CONT, consolidated in the unit's Executive Management Report, comparing 2024 (first year under the method) with 2023. The report shows an "18.2% in the aggregate" reduction; output per case remained stable (2.78 vs 2.77).

"The unit’s central indicator, average processing time, fell from 17 days, 22 hours, and 11 minutes (17.92 days) in 2023 to 14 days, 15 hours, and 44 minutes (14.66 days) in 2024. The reduction was 3.26 days per case, or 18.2% in the aggregate."

Discussion¶

Related Claims¶

• In UCI/SEDET's first year under the method (2025), average processing time fell 50%, cases processed rose 31%, documents 67%, technical reports 85%, and dispatches 60% — related
• SES/CONT's time gains translated into an estimated released capacity of about 352 additional cases (+22.3%) and cleared backlog, and output shifted toward preventive work (Advisory Notices +664%) — related