From the Emergency Services edition of November 6, 2026
South Florida flood night: city AI voice agent routed 227 of 298 resident calls to the EOC in a median 26 seconds
Unnamed South Florida city / Workforce Wave · Emergency management / municipal EOC public line · South Florida, United States
- Publisher
- Workforce Wave
- Original publication
- 2026-11-03
- Source retrieved
- 2026-11-07
- Event date
- 2026-11-02
What happened
Workforce Wave's Nov. 3, 2026 case study reports that during a South Florida local emergency after Nov. 1 flooding, an existing municipal AI voice agent on the city's public line was reconfigured the same afternoon to route flood/EOC intents. From Nov. 2 ~noon to Nov. 3 10:42 PM ET (~35 hours), the agent answered 298 resident calls and connected 227 to the EOC phone bank (median 26 seconds greeting-to-handoff; ~22 seconds for yes-to-greeting callers). Ten EOC connections occurred midnight–7 AM. Other outcomes: 21 answered/redirected without transfer, 31 hung up under 20 seconds, plus small transfers to city/water/admin lines. Life-threatening callers were told to hang up and dial 911; the agent never transfers those. City, partner, and residents are unnamed. Platform reliability claims (99.69% of 36,080 agency calls YTD) are vendor-reported.
Why it matters
Direct U.S. emergency-management signal: shows how a non-emergency municipal voice agent can be pressed into EOC surge routing without changing the city's PBX, with measured handoff latency—useful for coastal and flood-prone SLED agencies, with strong vendor-claim caveats.
Evidence and measured results
Case study dated November 3, 2026 by Colin P. Highland; disposition table for 298 calls; states 'Connected' means network handoff to the EOC line; reliability figures cover Jan 1–Oct 13, 2026.
Limitations and uncertainty
Vendor-authored case study; city unnamed; no independent audit; no clinical or property-loss outcomes; 'connected' is telephony handoff, not confirmed human answer or resolved need; bilingual example is anecdotal.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-11-07; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- EOCs need a public phone front door that stays up and routes flood callers quickly when city hall lines and staff are disrupted.
- Stakeholders
- Emergency management director
- EOC phone-bank lead
- City communications
- 911/PSAP coordinator
- Municipal IT/telephony
- Elected officials
- Discovery questions
- Is there already an AI or IVR on the main city line that can gain an EOC route without a new number?
- Who authorizes same-day greeting and intent changes during a declared emergency?
- How are life-threat callers forced to 911 rather than EOC transfer?
- Will post-event call dispositions be publishable?
- Value hypothesis
- If a city already runs a public voice agent, a pre-approved EOC pathway may cut median handoff toward the tens of seconds under flood surge—operational access value only, as claimed here.
- Potential engagement
- Bounded readiness workshop: design and tabletop an emergency voice-agent pathway with metrics matching this disposition table.
- Unsupported claims
- Guaranteed 26-second EOC connect in any city
- Property or life-safety outcomes from the flood
- That the unnamed city's results generalize
- Replacement of 911 by the voice agent
Pre-sales engineering
Role takeaway
- Fit
- Fits municipal voice agents / contact-center AI that already route non-emergency calls and can add disaster intent trees.
- Architecture and integration
- Public DID → cloud voice agent → conditional transfer to EOC ACD; ANI capture; no life-threat transfer; parallel normal department routes.
- Prerequisites
- Existing public-line agent with changeable prompts
- EOC trunk/ACD capacity for overnight surge
- Intent lexicon for flood/evac/shelter
- Logged configuration versions
- Constraints
- Unnamed city and partner stack
- Vendor self-report of connect rates
- Resident cellular outages still limit reachability
- Security
- Treat ANI, recordings, and transcripts as sensitive municipal emergency records; restrict vendor access and define retention.
- Proof of value
- Reproduce disposition table in a flood drill
- Measure median greeting-to-EOC connect with EOC answer confirmation
- Verify life-threat utterances never transfer
Delivery
Role takeaway
- Work
- Build and exercise a flip-switch emergency pathway with EOC staffing plan and after-action metrics.
- Dependencies
- Voice-agent operator
- EOC supervisors
- City PIO for published number messaging
- PSAP liaison for 911 boundary
- Ownership
- Emergency management owns pathway activation; telephony/IT owns config; EOC floor owns answering capacity.
- Skills and adoption
- EOC call-takers need scripts for AI-handed callers; PIO needs clear public wording that the AI is a router not a 911 substitute.
- Governance checkpoints
- Pre-season pathway approval
- Activation logged under emergency declaration
- Post-event disposition and misroute review
- Acceptance criteria
- Median handoff at or below agreed SLA in drill
- Zero life-threat transfers in test set
- Published after-action with disposition counts
- Risks
- Public confusion if news lists AI number as 'EOC'
- EOC queue overload from easy connects
- Vendor narrative outrunning independent verification
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Keep emergency pathways as configurable routes on an existing public voice agent (greeting question + flood lexicon → EOC) rather than publishing a new number mid-disaster; preserve hard 911 hang-up guidance with no AI transfer of life-threat calls.
Governance
Who approves, reviews and stays accountable for outcomes?
EOC surge configs need pre-approved playbooks, logged change control, and post-incident review of misroutes—especially when the public line is branded as the EOC number in local news.
Security and privacy
What data, permissions and controls need testing?
Automatic capture of caller phone numbers without name intake speeds handoff but still creates PII/call-detail records that must follow municipal retention and disclosure rules; cloud hosting outside the disaster zone is claimed as resilience.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Spanish-then-English example suggests language switching helps some callers, but equity for disability, low-literacy, and cell-outage populations is not measured; overnight EOC staffing remains the bottleneck the agent feeds.
Procurement
What should contracts, pricing and exit terms secure?
Prefer contracts that already include emergency-pathway change rights, audit logs, and post-incident data exports rather than buying a new flood chatbot mid-event.
Operating model
Which teams own the service once it runs?
Treat the public voice agent as dual-use: day-to-day 311-like routing plus a declared-emergency EOC front door with human answer capacity sized for the connect rate.
Publication history
- 2026-11-06Emergency Services · Issue 614 resources
Stable resource ID: workforce-wave-flood-night-eoc-ai-voice-agent-20261103