On September 30, 2024, days after Hurricane Helene reached western North Carolina, the Federal Communications Commission reported that 54 percent of cell sites in the affected part of the state were out of service. A September 2026 study of those outages adds a detail worth sitting with: the towers were mostly still standing. Physical damage to cell sites accounted for about 1 percent of outage time across six states. Most of the rest came from lost power and cut backhaul lines.
No agency can harden its way out of a severed fiber line in the next county. If accounting for your people depends on the network, it will fail exactly when you need it most.
The question that can't wait for a signal
Every organization that deploys people eventually has to answer one question under pressure: who is where, and is everyone accounted for?
An incident commander asks it about a disaster workforce spread across staging areas, shelters, and field sites. A public health lead asks it about staff who deployed, moved, and rotated out. A continuity manager asks it after a building is evacuated. A state emergency manager asks it as mutual aid crews arrive from other states.
When the network drops, the answer falls back to radio calls, paper sign-in sheets, and spreadsheets rebuilt at the end of a shift. By the time the roster is current, the decisions it should have informed are already made.
Defense learned this first. After Hurricanes Katrina and Rita, the Department of Defense had so much difficulty accounting for its own people that it directed each of the services to adopt an automated accountability system. The tactical edge, where networks are least reliable, has been the proving ground ever since.
Built for the edge, not adapted to it
Q Guide started in that defense work, as personnel accountability for military operations in denied and disconnected environments. The same application now serves civilian response as part of DMI's Q Suite.
Q Guide is one browser-based application that runs on the phone, tablet, or laptop a person already carries. There is nothing to install, no app store, and no device enrollment. It works whether the connection is full, intermittent, gone, or deliberately air-gapped. When the network drops, Q Guide keeps working:
- One locked roster that holds everyone on the operation, including partner agencies, contractors, and volunteers.
- Check-in and check-out at every checkpoint, verified against that roster, with discrepancies flagged on the spot.
- Credentials that verify offline, issued as QR codes, barcodes, or NFC when someone's ID was lost in the event. A person whose ID went into the debris field is still verifiable at the next checkpoint.
- Encrypted data on the device, with every action signed and written to an audit record.
- Device-to-device transfer through a sequence of QR codes, screen to camera, when there is no network at all.
When the link returns, updates sync by priority and structured accountability data flows back to the systems that depend on it. Offline operation is not a mode added later. Every feature was built on the assumption that the network would fail.
AI that still answers during a blackout
DMI Q Guide also brings AI to the accountability record. Instead of building a report, a supervisor asks a plain-language question, such as who on this roster hasn't checked in since the last movement, and gets an answer grounded in the accountability data.
Two design choices make that workable in government. First, it works offline. A compact language model runs inside the browser against the encrypted data on the device, so the answer is there with no cloud connection. Second, it is governed. Every request is screened for sensitive data before anything is processed, and that screening decides where the request is allowed to run: inside the agency's boundary, on an approved government cloud endpoint, or not at all. Sensitive fields are tokenized on the way out and restored in the answer, so the model provider never sees them. Every AI tool inherits the user's role-based access, and every call is logged.
That is the bar public sector leaders are setting. State CIOs ranked AI their top priority for 2026, ending cybersecurity's 12-year run at number one, and framed the work around governance, security, and privacy.
The same application, very different worst days
- Defense. Force posture, movement and checkpoint accountability, casualty workflow, and accountability data published for NGC2 and the Army's Edge Data Mesh, on IL5 and IL6 operating paths.
- Federal civilian. Incident workforce deployment and daily accountability, public health emergency deployments, and continuity of operations.
- State and local. Emergency management and mutual aid, and campus evacuation or shelter-in-place accountability.
One codebase runs at IL5 for defense missions and IL2 for civilian response, with no re-architecture between them. Q Guide doesn't ask agencies to replace the workforce or workflow platforms they already run. It adds the field accountability layer those platforms were not built to provide.
Operated by the team that built it
DMI doesn't hand Q Guide over and walk away. We deploy it, run it, and integrate it, backed by a 24×7 service desk, in the cloud on AWS, on premises, or fully air-gapped. Pricing follows mission scope rather than named users, so bringing on surge staff doesn't mean buying more seats. Q Guide is available through AWS Marketplace and through federal contract vehicles including GSA MAS, Alliant 2, Army ITES-3S, and NASA SEWP.
The next major disaster will take networks down again. The question is whether your accountability goes down with them.
SEE Q GUIDE WORK WHERE YOUR MISSION RUNS
Request a briefing at dminc.com/q-guide, watch the Q Guide film, or download the Q Guide fact sheet.
