Eight weeks. Three phases. Every DC CAP team. Here is what the program covers, the goals it sets, and how we know it is working.
Every hour Claude saves on drafts, data pulls, and formatting is an hour back for the 800+ scholars we serve. This program teaches every DC CAP staff member how to take that hour back — responsibly. Ten senior leaders across six units tested the model from April to June 2026. Now it comes to every team, funded by a $600K grant from the KPMG AI Impact Initiative.
Advance all participants toward regular AI use in daily workflows, measured by pre/post assessment across five fluency dimensions. 70%+ demonstrate three or more observable fluency behaviors by Day 60.
Create conditions where participants experiment openly, share failures productively, and build confidence through practice. Weekly confidence scores trend upward. Facilitation transitions from leader-led to champion-facilitated.
Build a peer learning network across units. Each unit produces at least one champion candidate. Participants teach techniques to peers.
Document measurable time savings and workflow improvements. Each participant captures at least one before/after comparison. Minimum two workflow redesigns per unit.
Move beyond doing the same things faster toward building fundamentally new capabilities. At least one unit demonstrates an AI-enabled capability that could not exist at current quality or scale without it.
All participants complete governance orientation and demonstrate responsible data classification. Zero Tier 1 or Tier 2 data incidents. AI-assisted outputs carry clear human ownership and review checkpoints.
Full goal detail and measurement framework: KPI Framework | Governance goals: AI Governance Framework
100% of participants have attempted Claude on a real work task
Each participant has identified at least one workflow where Claude saves time or improves quality
Measurable fluency growth on pre/post assessment; 3+ reusable workflow templates documented; each unit's capstone shipped