November 17, 2026
providence-leverages-ai-guided-coaching-to-support-workforce-transitions-through-rapid-deployment-of-my-conversation-companion

Providence, one of the largest non-profit health systems in the United States, has successfully implemented a proprietary artificial intelligence tool designed to coach its leadership through complex organizational changes. Known as My Conversation Companion, the AI-guided agent was deployed to 1,300 core leaders within a 10-day window to support a significant workforce transition. While the speed of the rollout appeared to be a reaction to immediate organizational needs, it was the culmination of a multi-year development cycle characterized by iterative testing, technical adaptation, and strategic patience. The initiative marks a significant milestone in how healthcare organizations utilize generative AI to manage human capital and internal culture during periods of volatility.

The Evolution of My Conversation Companion

The development of My Conversation Companion began in early 2024, coinciding with the global surge in generative AI adoption. While many organizations focused on utilizing large language models (LLMs) for administrative automation or clinical documentation, the development team at Providence identified a gap in leadership development. The objective was to create a "safe space" where leaders could practice high-stakes interpersonal scenarios—such as delivering difficult feedback or influencing stakeholders without direct authority—using an AI-guided coach before engaging in real-world interactions.

The project followed a non-linear path, evolving through five distinct environmental shifts. Initially, the concept was developed in what internal stakeholders referred to as a "greenhouse"—an isolated environment using personal ChatGPT accounts to test basic conversational patterns. This allowed the developers to refine the AI’s logic without risking proprietary data or violating organizational security protocols. By keeping the initial experiments generic, the team focused on the "pattern" of human conversation rather than a specific product, ensuring that no Protected Health Information (PHI) or Personally Identifiable Information (PII) was ever introduced into external systems.

As Providence’s internal infrastructure matured, the tool moved through several enterprise platforms. Later in 2024, it was migrated to ProvidenceChat, an internal, secure instance of an AI chatbot. However, this version faced scalability issues due to the lack of centralized delivery mechanisms. By early 2025, the sunsetting of ProvidenceChat led to a third iteration built on Microsoft Copilot. While the user experience improved, the tool remained limited by the developing governance frameworks surrounding enterprise-wide AI deployment. The final technical breakthrough occurred in late 2025 with the adoption of Copilot Studio, which provided the necessary infrastructure for large-scale deployment and integrated governance.

Strategic Readiness and the Ten-Day Sprint

The rapid deployment of the tool in early 2026 was triggered by a workforce transition designed to bring teams back into the office and strengthen the health system’s organizational culture. This transition required 1,300 core leaders to navigate sensitive conversations with their teams regarding work arrangements and cultural expectations. The urgency of the situation required a support mechanism that could be scaled instantly—a demand that My Conversation Companion was finally positioned to meet.

Despite the 10-day timeline for the final deployment, the project had undergone nearly two and a half years of preparation. This "strategic readiness" allowed the team to repurpose the existing AI patterns to align with the specific nuances of the workforce transition. The governance process, which typically serves as a bottleneck for new technology in healthcare, was expedited because the tool had already been vetted through previous iterations.

During the period leading up to the launch, the project entered what developers called "the pause." This was a phase where the technical build was complete, but a high-priority business case had not yet emerged. Rather than remaining idle, the team used this time to refine messaging, build supportive SharePoint infrastructure, and engage with various stakeholders across the enterprise. This proactive preparation ensured that when the executive mandate arrived, the delivery mechanism was as polished as the AI logic itself.

Data-Driven Results and User Sentiment

The success of the deployment was measured through both quantitative metrics and qualitative feedback. Among the leaders who engaged deeply with the tool, My Conversation Companion achieved a Net Promoter Score (NPS) of +54. In the context of internal enterprise software, a score above +50 is considered "excellent," indicating high user satisfaction and perceived value.

The feedback from the 1,300 core leaders highlighted a significant shift in leadership confidence. Key findings from the deployment included:

  • Confidence Building: Users reported that the AI’s ability to simulate emotional responses allowed them to move from a state of "reaction" to "thoughtful action."
  • Human-Centric Feedback: Participants praised the "eye-opening" nature of the coaching feedback, which pointed out blind spots in their communication styles.
  • Technical Friction: Negative feedback was primarily focused on technical delivery—such as the speed of the interface or the desire for voice-to-voice interaction—rather than the quality of the AI’s coaching.
  • Demand for Expansion: Several leaders requested that the tool be made available to their entire teams, suggesting that the utility of AI coaching extends beyond executive leadership.

The deployment was deliberately capped at 1,300 leaders to ensure a controlled environment. This responsible approach to innovation allowed Providence to monitor the system’s performance and gather high-quality data before considering a broader rollout to the organization’s 120,000+ caregivers.

Executive Sponsorship and Organizational Impact

The success of the initiative was heavily dependent on executive sponsorship. Sandra Murray, a key executive sponsor at Providence, emphasized that the tool’s value was not in its novelty, but in its ability to solve a persistent business challenge: providing meaningful, scalable skill practice for high-stakes conversations.

"I used the tool myself to prepare for conversations I knew would be difficult," Murray stated. "It gave me a safe place to anticipate emotions, consider perspectives, and practice navigating those moments with compassion. The risk of doing nothing was greater than the risk of trying something new."

Murray’s perspective underscores a critical shift in how healthcare leadership views AI. Rather than seeing it solely as a tool for efficiency, it is being recognized as a vital instrument for maintaining the organization’s "Mission and Promise"—specifically the core tenets of "Know Me, Care for Me, Ease My Way." By prioritizing the "person first" in the tool’s messaging, the project team was able to gain buy-in that a purely technology-focused pitch might have failed to secure.

Analysis: The Five-Step Roadmap for AI Innovation

The Providence experience offers a blueprint for other large organizations looking to integrate generative AI into their talent and learning operations. The project team identified a five-step journey that moves beyond the initial prototype:

  1. The Prototype: Developing the initial AI logic in a controlled environment.
  2. The Enterprise Platform: Moving the concept into a sanctioned, secure corporate environment.
  3. The Governance Path: Earning executive sponsorship and navigating legal, ethical, and security reviews.
  4. The Business Case: Identifying the specific "moment of need" where the tool provides maximum value.
  5. The Deployment and Scaling: Managing the messaging and delivery to the end-user.

This roadmap highlights that the technical build is often the simplest part of the process. The greater challenge lies in the "reframe"—transforming a learning tool into a strategic instrument that provides executives with real-time signals regarding organizational sentiment and leadership readiness.

Broader Implications for the Healthcare Industry

Providence’s successful use of My Conversation Companion suggests that the future of AI in healthcare will be increasingly focused on the "human core" of the industry. As health systems face ongoing challenges related to workforce retention, burnout, and cultural shifts, the ability to provide on-demand coaching at scale becomes a competitive advantage.

Furthermore, the project demonstrates that "shadow innovation"—starting small and outside the main enterprise environment—can be done responsibly if strict data guardrails are maintained. This approach allows for rapid iteration that would be impossible within the rigid structures of a traditional healthcare IT department.

As Providence prepares for the next phase of the project, the focus will shift toward the "signals" the tool generates. By analyzing the aggregate, anonymized data from leadership interactions, executives can gain insights into the specific anxieties and challenges facing their workforce in real-time. This predictive capability could transform how healthcare organizations approach change management, moving from a reactive posture to one of proactive support.

The story of My Conversation Companion is ultimately a testament to the importance of preparation. In an industry defined by uncertainty, Providence has demonstrated that the most effective way to handle a crisis is to have the solution already built, tested, and waiting for its moment to arrive.