Pilot 100: What the Pilot Must Prove

What Pilot 100 is designed to prove

100 residents. Existing Alabama resources. One coordinated pathway. Measurable results.

Pilot 100 is a proposed bounded implementation and integration test. It asks whether coordinated assessment, service access, documented responsibility and community handoffs add value beyond existing practice. It does not promise a statewide rollout or establish causal effectiveness simply because 100 people participate.

• Connect existing services without creating competing enrollment, referral or reporting processes.
• Create an individualized pathway with named responsibility for each unmet need and handoff.
• Improve staff access to usable information while measuring workload and duplicate entry.
• Test approved technology with human review and independent downtime procedures.
• Track safety, treatment continuity, education, employment and reentry outcomes.
• Report failures, unmet needs and unintended effects, including reasons a component should not expand.

Conditions before enrollment

• Secure facility authorization, a sponsoring operational owner and written partner responsibilities. Participation cannot substitute for lawful release or parole procedures.
• Verify available programs, eligibility, waiting lists, staffing and costs at the selected site. Align with RESTART and existing referral networks where geography and eligibility permit.
• Define the cohort, selection criteria, exclusions and safe separation plans. Prioritize the proposed near-release and parole-preparation tracks without treating release proximity as proof of low risk.
• Maintain separate treatment and security tracks based on assessed needs and credible threat information. A completion certificate or an AI score cannot override unresolved safety concerns.
• Approve a data-sharing plan, access rules, correction process, evaluation plan and downtime drills before handling live resident data.
• Budget personnel, training, maintenance, support, evaluation and community services—not only devices or software. Name the funding source before committing expenditures.

Start small, then earn expansion

Start with synthetic data for demonstrations. No public demo app is approved for production resident, clinical, victim or security records.

Publish the evaluation rules before launch

The evaluator and operational owner must define baselines, numerators, denominators, missing-data handling, follow-up windows and success thresholds before enrollment. Report who was excluded, who left the pilot and who could not be reached. Where feasible, use a comparable cohort and explain selection bias; avoid attributing every change to ALPRP. A pilot of 100 may establish feasibility but cannot guarantee precise estimates for rare safety events or statewide recidivism effects.

Continue, redesign or stop

Expansion requires evidence of added benefit, feasible costs, acceptable staff burden, reliable delivery and an authorized governance review. Suspend an affected technology or workflow after a serious safety or privacy failure, unreliable identification, failed downtime drill or persistent unresolved data errors. Preserve care and essential services during any suspension. Publish de-identified findings and the corrective action before reconsidering expansion.

Specific operating thresholds and review intervals remain to be agreed before launch. ALPRP will not label the pilot successful by enrollment totals alone.

Current plan and review timing

Pilot 100 Master Plan v5.2 describes a five-year demonstration with an evidence-based continuation decision by the end of Year 2. This overview clarifies proposed implementation safeguards; it does not replace or revise the controlling master plan.