Frontline workers are roughly 80% of the global workforce, and the part of it technology has most consistently passed over. Listen to the market and you'd think otherwise. Nearly half of S&P 500 earnings calls now feature AI prominently. For public-sector leaders that's the elephant in the room, loud at the executive level, largely absent from the floor where the work of government service gets done. The ambition lives at the top. The work lives on the floor. Most programs stall in the distance between them.

This roundtable brings public-sector operations, IT, and informatics leaders together to compare how they close that gap: learning what the work actually requires before committing to a platform, what changes when AI runs on a handheld instead of in a browser, when narrow domain-specific agents beat one large general model, and what earns a skeptical worker's trust in the first weeks.

Attendees leave with a sharper read on where their own frontline AI efforts are most likely to break, and a few practical moves for turning executive ambition into something the floor will actually use.

BWG Connect, Insight Public Sector & Zebra Technologies invite you to join an interactive event with industry peers to review & discuss practical insights across the digital space.

As always, there will be no sales pitch and all attendees are welcome to listen and learn.

• Most AI investment is decided in the boardroom, yet the people who would have to use it every day rarely shape what gets built, and that distance is where adoption quietly dies.
  • How leaders find out what frontline work actually requires before committing to a platform
  • Where the strategy-to-deployment handoff tends to break in public-sector operations
• Frontline environments punish the assumptions that work fine for knowledge workers, because the people on the floor are moving, time-pressured, and can't stop to wrestle with software.
  • What changes when AI has to run on a handheld or wearable at the edge rather than in a browser
  • The tradeoffs of on-device, token-less AI versus cloud-dependent tools when connectivity and cost are real constraints
• Leaders are discovering that one large, all-knowing AI model is often the wrong tool for a specific operational job.
  • When a set of narrow, domain-specific agents beats a single general model in practice
  • How teams decide which frontline tasks are worth purpose-building an agent for
• Even capable AI stalls on the floor when workers don't trust it or when it adds steps instead of removing them.
  • What earns frontline adoption in the first two weeks versus what gets quietly abandoned
  • How Insight and Zebra approach the integration and change-management work that determines whether a rollout sticks
• Most frontline AI programs can't clearly answer whether they actually improved operations, which makes the next round of investment a guess.
  • The operational metrics that separate a real productivity gain from a demo that impressed leadership
  • How peers are building the measurement case that unlocks funding to scale