The prompt asked for an approach to the first ninety days of stakeholder engagement. This is that approach in full, including who is in each session and what each phase has to produce before the next one starts.
Three commitments that shape everything below
- No procurement in the first ninety days. If we lead with tools we will have bought something before we know what we need.
- Labor partners and families in week two, not week ten. The sequencing is the design.
- Existing staff time and existing convenings. No new headcount is proposed. Board Goal 4 is to maximize human and operational resources, and the county's fiscal environment does not permit anything else.
Phase one, days 1 to 30: Reflect and Learn
The purpose of phase one is to be able to say, with evidence, where this organization actually is. Nothing in this phase is a decision.
Week 1
- Charter confirmed with the Senior Leadership Team: scope, authority, what the body does not do, and the two-year staggered-term structure.
- Seats confirmed across the seven named divisions, plus Business Services and the Data and Impact Center of Excellence.
- Opening communication in English and Spanish, so the first thing anyone hears about this body is not a rumor.
Week 2
- Labor partners. Not a briefing. A working session on what would have to be true for members to trust an AI initiative, and what must be negotiated rather than announced.
- Families. Using the parent advisory structures that already exist, interpreted, at times families can actually attend.
- Students. Through the student board members and the monthly student experience panels SDCOE already runs.
Weeks 3 and 4
- Listening sessions in all nine divisions, including classified staff, who use these tools and are almost never asked.
- An honest inventory of the AI already in use at SDCOE. Not an audit for discipline. An inventory, because the tools are in use whether or not they were authorized, and a plan built on a fiction is worthless.
- Readiness survey against the three levels of the divide: connection and devices, recurring cost, and educator capability.
- A short scan of what districts are already doing, so we build on the ones that are ahead instead of duplicating them.
Phase one deliverable
Readiness Baseline. One document that states where SDCOE and the county stand on all three levels, what AI is already in use, and what staff, labor, families, and students said. Everything in phases two and three is measured against it.
Phase two, days 31 to 60: Design and Align
The purpose of phase two is that people build the thing they will have to live under. This is co-creation as SDCOE defines it, which is centering those most affected as design partners rather than as an audience.
- Co-design studios, not briefings. One per division. Each produces two artifacts: the three tasks in that division where AI would help most, and the three places it must never be used.
- Draft guardrails built on the CDE model policy released in June 2026, adapted for county-office context. We do not write our own from scratch and we do not duplicate CDE. We localize.
- Role-based capability map. The 201 and 301 material, defined per role rather than generically, because a program specialist, an HR analyst, and a classroom teacher do not need the same thing.
- Industry roundtable. Employers in this county, describing what changed in their work in the last eighteen months. This informs pathways, and it informs what we tell students is coming.
- Accessibility and language review of every artifact before it goes out, to Section 508 and WCAG, in English and Spanish.
Phase two deliverable
Draft Guidance and Capability Map. A guardrails document districts can adopt or adapt, and a role-based map of what 201 and 301 capability actually means at SDCOE.
Phase three, days 61 to 90: Develop and Refine
The purpose of phase three is proof. Small, real, and measured, including the parts that do not work.
- Three pilots in three different divisions. Deliberately small, deliberately unglamorous, and chosen because they touch a task someone does every week.
- Measured against the phase-one baseline, not against enthusiasm.
- Publish what did not work. A task force that only reports wins stops being believed, and the credibility of this body is the asset the whole program runs on.
- Board-ready recommendation to the Senior Leadership Team, in language that can go to the County Board without translation.
- Year-one plan with named owners, sequencing, and honest cost, built on existing staff time.
Phase three deliverable
Board-Ready Recommendation. Plus the one page a superintendent can actually use, which is the real test of whether ninety days were well spent.
Why these phase names
Reflect and Learn, Design and Align, and Develop and Refine are the three phases of SDCOE's own 2026-27 LCAP Development Series. That is deliberate. This ninety-day window sits inside the LCAP development cycle, which means that if the task force delivers on time, districts can write AI access, AI literacy, and professional learning into three-year plans with funding attached, rather than bolting them on afterward. If the window is missed, the next clean opportunity is three years out.
How success is measured
No new scorecard. Four outcomes SDCOE already reports:
| Outcome | What it measures | Why this body moves it |
|---|---|---|
| Desired Outcome 1 | Staff who strongly agree they belong at SDCOE and that it builds trust | How this work is done determines whether it builds trust or spends it |
| Desired Outcome 2 | Staff who can access professional growth, co-creation, and cross-division collaboration | This is nearly a definition of the task force. It is the primary measure. |
| Desired Outcome 3 | Families and partners who say SDCOE co-creates with them | Families are engaged in week two, and the guidance is built with them |
| Desired Outcome 7 | County LEAs improving on the College and Career Indicator | The industry roundtable feeds pathway design directly |
What this body does not do
- It does not purchase. It advises.
- It does not police. Districts are independent, and a county office that behaves like a regulator loses the influence that makes it useful.
- It does not duplicate CDE. It localizes CDE for a county where one district is 46 percent English learners and another is 3 percent.
- It does not survive on one person. Staggered terms exist because an initiative that cannot survive a change of leadership is not an initiative.