Deciding Where AI Can Be Trusted in a 911 Call Center

Situation

Next Generation 9-1-1 raised a question nobody had a settled answer to: where can AI be safely deployed in a call center, where a wrong model output costs a life? Emergency services providers also needed to modernize their technology stack. The usual failure mode in that world: a vendor shows up with a platform pitch before anyone's documented what's actually broken on the ground. Investment decisions were getting made off vendor demos, not operational reality.

What I Built

• Determined where AI can be safely deployed in 9-1-1 call centers, and built the governance criteria for making that call.
• Ran primary field research across vendors and live deployment sites, separating the AI capabilities that held up in operational use from the ones that did not, including drone-as-first-responder providers.
• Authored two 50-page operational assessments for emergency services providers, delivering vendor evaluation criteria, competitive analysis, and prioritized modernization roadmaps that drove pilot and procurement decisions and fed national guidance on AI adoption in emergency communication centers.
• Ran six customer discovery engagements with operations leadership, translating frontline pain points directly into prioritized product requirements and integration recommendations, not a vendor's feature list.
• Convened 50 practitioners, vendors, and policy stakeholders in Washington, DC, and delivered 30 executive briefings, shaping investment prioritization, vendor selection, and policy direction for AI adoption in emergency response.
• Facilitated cross-organizational strategy and governance sessions, producing executable roadmaps that covered pilot, scale, and steady-state operations, not just the exciting first phase.

The Hardest Problem

Getting operations leadership, the people who actually run emergency response, to trust that a technology assessment came from their reality, not a consultant's template. That trust got earned in the discovery engagements themselves, with questions specific enough to prove the work started from operational pain and not a sales deck.

Outcome

Two comprehensive assessments gave emergency services providers a procurement-ready path forward, grounded in documented operational needs, not vendor promises. The roadmaps covered the full lifecycle from pilot to steady-state operations. Most modernization plans skip that part.

Skills Demonstrated

AI governance criteria, operational assessment and technology strategy, customer discovery at the executive level, vendor evaluation frameworks, cross-organizational governance facilitation, public-sector modernization planning.