Skip to content

for-internal-teams.md

Latest commit

 

History

History
97 lines (62 loc) · 4.99 KB

for-internal-teams.md

File metadata and controls

97 lines (62 loc) · 4.99 KB
Raw

Adaptation — Internal Teams (Community Pods, AI Champions, Learning Cohorts)

For running this curriculum inside your own organization — community-of-practice pods, AI Champions programs, departmental learning cohorts.

What's different about internal cohorts

  • Audience is consistent. You'll see the same people across multiple sessions. Build for arc, not for one-shot.
  • Scenarios are easy to source. Real internal workflows are everywhere. Use them.
  • Peer pressure works. Cohort dynamics — visible progress, peer demos, shared backlog — drive adoption harder than facilitator energy.
  • Time is scarce but recurring. Don't fight for a 4-hour event. Win 60 minutes every two weeks for 6 months.

The community pod format (recommended)

A 6-month commitment with a small cohort (8–15 people) and a recurring 60-minute biweekly session.

Cadence Session
Week 1 Track 0 L100 — Responsible AI
Week 3 Track 1 L100 — Copilot CLI
Week 5 Track 3 L100 — Prompt Engineering
Week 7 Track 2 L100 — Building Agents
Week 9 Track 5 L100 — Memory & Tokens
Week 11 Track 6 Topic 2 L100 — Workflow Design
Week 13–24 Cohort picks L200/L300 deep-dives based on what they actually need

Why this works: the off-weeks (between sessions) are where adoption happens. Each session ends with a homework: "try this in your real work this week, demo it next session." Peer accountability does the rest.

The AI Champions variant

For a Champions program (peer mentors who get a small time allocation to support adoption):

  • Champions go through the full curriculum L100 + L200 as a cohort over ~3 months
  • Champions then run L100 sessions for their own teams as part of the program
  • Monthly champion-only debriefs: what's working, what's not, what's missing

The curriculum becomes the shared vocabulary of the Champions program. New champions onboard via the same path. The program scales without re-inventing materials each time.

Sourcing internal scenarios

The single biggest advantage of internal cohorts is access to real workflows. Don't waste it.

Three reliable scenario sources:

  1. Cohort intake survey. Ask every cohort member to describe one workflow they spend more than 2 hours/week on. Use the most common one as the recurring scenario across sessions.
  2. Their managers. Ask managers what they wish their team did more efficiently. Often surfaces workflows the cohort members themselves overlooked.
  3. Recent incidents or escalations. Real failures are great teaching material. With permission and anonymization, work through "what would AI have changed here?" in the relevant session.

See scenarios/README.md for guidance on writing reusable scenarios.

Cohort metrics that matter

Don't measure attendance. Measure:

Metric How to capture
Pre-work completion rate Microsoft Learn completion API or honor-system survey
In-session activation "Did you try the technique live in this session?" 1-question poll
30-day usage Light-touch follow-up: "Did you use this in real work this week?"
90-day workflow change Specific: "Has any recurring workflow in your role changed because of this curriculum?"
Cohort-level demos Number of cohort members who demoed their own work to peers

The last one — peer demos — is the leading indicator of program health. A cohort where members are demoing to each other unprompted is succeeding. A cohort where the facilitator demos every time is not.

What success looks like for internal cohorts

  • 30-day: 70%+ of the cohort tried at least one technique in real work
  • 90-day: at least 3 cohort members have made a measurable workflow change they can articulate
  • End of program: the cohort has produced 3+ scenarios you can publish to the scenarios/ library
  • Long-term: alumni of the cohort become facilitators of the next cohort. That's the loop.

Common pitfalls

  1. Recruiting the cohort too broadly. A cohort of 50 is a webinar. A cohort of 8–15 is a learning community. Pick the smaller number.
  2. Letting attendance lapse without consequence. If cohort members can skip 3 of 6 sessions, they will. Set an expectation up front: "Attend 5 of 6, do the homework, demo at least once. Otherwise let someone else have the seat."
  3. No alumni path. Cohorts that end without an "ok, what now?" lose 80% of their long-tail value. Have a Champions program, advanced cohort, or contribution path ready to offer to high-engagement alumni.

Light-touch variant (60 minutes total)

Don't have 6 months? Run a single Track 0 + Track 1 L100 combo as a one-off lunch-and-learn. Manage expectations: this is exposure, not enablement. Offer the full cohort program as the next step for anyone who wants more.

This is also the right format for a "do we want to do the full cohort?" pitch session to leadership.