Quick Links
The Veterans Health Administration (VHA) Innovation Ecosystem wanted to get practical about AI: where could it genuinely help Veterans — especially in mental health and suicide prevention — and what would it take to do it responsibly?
The challenge was that the right expertise lived in very different places: clinical decision-making, medical research, data science, policy, executive leadership, and frontline operations. Bringing those perspectives into one (virtual) room and turning them into a shared, actionable plan called for careful facilitation, not another presentation.
We designed and facilitated a two-day AI & Mental Health workshop built around a single purpose: bringing cross-disciplinary champions together in service of veteran-centric AI innovation.
what we did
Designed and facilitated a two-day virtual AI & Mental Health workshop for the VHA Innovation Ecosystem, including interactive World Café sessions to build shared understanding of the problem space and AI Design Sprints across seven Veteran-centric use cases
Used a custom set of AI Capabilities Cards so non-technical experts could map real AI capabilities to their use cases
Captured insights, questions, and concrete next steps for each use case, including suicide-risk prediction and a Veteran-facing wellness chatbot
how we did it
A cross-disciplinary group, by design
The summit deliberately mixed clinical, research, policy, executive, and operational voices — the people who rarely get in one room but who all have to agree if AI is ever going to reach Veterans safely. Getting them to genuinely hear one another was the first job.
Throughout the summit, every team was pushed to name the hard questions — consent, privacy, equity, and Veteran trust — right alongside the opportunities, so the ideas that emerged were responsible, not just clever.
Day 1 — World Café: building shared understanding
We opened with a World Café: three 20-minute rounds in small groups, reshuffled each time, moving through escalating questions:
- “What’s important to you about this situation and why do you care?”
- “What challenges might come our way and how might we meet them?”
- “Imagine we were wildly successful — what bold steps might we take together?”
By the end, a room of specialists shared a common picture of the problem.
Day 2 — AI Design Sprint: from problem to prototype thinking
We split the group into seven pre-assigned use-case teams — Clinical Decision Making, Risk Management (Suicide Risk Prevention and Prediction), Documentation Support, Contact Center AI assistance, Veteran-Centric AI Support for Communications, Quality Assurance, and Population Health Monitoring — each with a dedicated facilitator. Using a deck of AI Capabilities Cards, teams broke down their use case, chose the AI capabilities that could actually help, and mapped a focused solution.
The outcome: momentum and a plan
Each use case left with concrete next steps — for example, using AI to segment risk and predict demand for suicide-prevention resources, and a Veteran-facing wellness chatbot that could help build safety plans and connect Veterans to care.
The solutions and process were all documented in a report which was then sent to leadership through the VA.
As a result of the momentum, 100% of attendees wanted another AI summit hosted in the near-future.
