Summary

wOjSGO70FPAdyluAQ0Z61p_FRf8jrLAr5BFDI72hUByohJhS6EkPSnbS72bE63SjOvts6-eCWkZOjmI1jRQR_VYqFyYssXy1w3qcKmZAhm5jW6sk9JlpLvI4-Zc_Yg==
Edit Back
Meeting Date
Unknown
Meeting Type
Call
Note Type
Summary
Attendees
Shawn Remick, Casey Cavanaugh

1. Meeting Summary

I met with Shawn Remick to get aligned on the current AI program direction and to establish Casey as the primary SawStop coordinator for these efforts. The main objective was to reduce the confusion caused by multiple parallel conversations and create a clearer operating model for policy, training, tool adoption, and future use case development.

Shawn walked through the current state of the AI initiative, which is centered on three primary workstreams: policy and training, targeted enablement through live training sessions, and broader alignment on the strategic direction for tool usage and business use cases. The policy is being revised to be less restrictive in tone and more practical in application, with a red light, yellow light, green light framework for data usage. I agree this is a better direction and is clearer than the earlier version.

On training, they are nearly finished with the current rollout at TTS/Festool and want to begin the same training for SawStop from mid-to-late April through May. The model is intentionally small group, live sessions with about 12 people per class to keep the sessions interactive and useful. Casey will help identify the SawStop employee population that should be invited, with the first pass focused on employees who spend most of their day at a computer rather than field, repair, or warehouse roles. Participation is optional, but leadership support will be important to drive adoption.

We also discussed the emerging use cases that are gaining traction. Shawn highlighted the work Brian Weiss is doing with a Copilot-based proprietary agent in service/marketplace operations, which is already showing strong value and may become one of the featured examples in the May collaborative. We also discussed the prior exploration of a NetSuite connection through ChatGPT. That path did not perform well, and the team now prefers to explore a Copilot-based approach instead because it provides better user-level access control and appears more scalable and cost effective.

Casey was supportive of the overall direction and agreed to help coordinate messaging, identify training participants, and reinforce the change management side with SawStop leadership. He also plans to attend one of the remaining live training sessions himself to better support the rollout.

2. Attendee List

  • Shawn Remick
  • Casey Cavanaugh

3. Action Items

  • [Casey Cavanaugh] Review the employee tracker and identify which SawStop employees should be offered AI training.
  • [Casey Cavanaugh] Mark the proposed SawStop training participants in the spreadsheet this week.
  • [Shawn Remick] Add Casey to the AI program management group and provide access to the spreadsheet and training materials.
  • [Shawn Remick] Create a SawStop-specific training proposal column in the tracker.
  • [Shawn Remick] Send Casey the training deck/materials for review.
  • [Shawn Remick] Forward Casey the invite for the remaining live training sessions.
  • [Casey Cavanaugh] Attend the March 19 live training session.
  • [Shawn Remick] Draft proposed SawStop training dates, likely on Fridays, for the April-May rollout.
  • [Casey Cavanaugh] Socialize the AI direction, policy updates, and use case concepts with SawStop stakeholders and leadership as needed.
  • [Shawn Remick / team] Recirculate the revised AI policy with updated tone and the red light, yellow light, green light data framework.
  • [Shawn Remick / Ronnie] Explore a Copilot-to-NetSuite approach for controlled query use cases, including leadership decision support and possible CRM/service data access scenarios.
  • [Casey Cavanaugh] Request a Copilot license after training.
  • [Casey Cavanaugh] Follow up with Ronnie on sending the draft meeting recording policy to Shawn and team for review.
  • [Shawn Remick / team] Review the meeting recording guidance and determine whether it should be incorporated into broader acceptable use policy updates.

4. Relevant Timelines

  • Mid-to-late April through May - Target window for SawStop AI training rollout.
  • Weekly cadence during rollout - Training sessions expected once per week, likely Fridays.
  • This week - Casey to complete the first pass of the SawStop participant list in the tracker.
  • March 19 - Casey plans to attend an existing live AI training session.
  • End of current TTS/Festool rollout - Two remaining training sessions before pausing and shifting focus to SawStop.
  • May - Team expects to highlight leading AI use cases, including Brian Weiss's service/marketplace operations example, during the collaborative.
  • Later this year - Goal to enable reporting on Copilot license utilization so unused licenses can be reclaimed.

5. Additional Notes

  • The AI policy is moving toward a more pragmatic governance model rather than a prohibitive one. The red light, yellow light, green light structure should make adoption easier while still protecting sensitive data.
  • The team is intentionally trying to balance enablement with cost discipline. They are not planning to fund every tool request and want to stay selective about where licenses and development effort go.
  • Training is optional by design, but the expectation is that leadership encouragement will materially improve participation and downstream tool adoption.
  • Current training completion rates are running around 31 percent of those offered, though this is likely being suppressed by AdaptX workload and competing priorities.
  • There is clear momentum around Copilot as the preferred enterprise platform, especially where access control and integration with internal systems matters.
  • Some functions, especially marketing, may continue to have legitimate use cases for tools like Gemini, Claude, or ChatGPT for public or creative work. The organization is trying to avoid over-policing those scenarios when company data risk is low.
  • There are already some employees carrying both Copilot and ChatGPT licenses. The longer-term model may evolve toward a limited multi-tool allowance, but no formal rule is in place yet.
  • Brian Weiss's work stood out as one of the strongest current examples of business-led AI adoption. It appears to be producing real time savings and positive customer outcomes.
  • Casey and I had previously discussed guidance around meeting recording. That draft has not yet made it to Shawn, so I need to make sure Ronnie resends it. The likely long-term home for that guidance is within acceptable use policy updates rather than as a standalone item.