Workshop infrastructure reliability checklist

This checklist is for facilitators who run medium/large RoomRadar sessions and need repeatable reliability under pressure. Use it as an operations runbook, not as a compliance form. In a hurry: assign one reliability owner, complete pre-flight checks, run fixed monitoring sweeps, and close with incident notes plus follow-up actions.

Pre-session checklist (30 to 10 minutes before start)

Room and mapping

• physical table labels match RoomRadar group names
• spare labels are ready if table count changes
• table map is visible to facilitator and co-facilitator

Devices

• one planned active phone per table
• one charged spare phone per facilitator zone
• spare phones are unlocked and pre-opened on join page
• charging/power plan exists for long sessions

Network and capture

• quick walk test in all table zones
• weak zones are marked for early monitoring
• one test sentence verified in each group stream
• no accidental duplicate active devices

Launch controls (first 5 minutes)

1. Run one room-wide readiness line table by table.
2. Confirm fresh lines from each active table.
3. Classify incidents as local or systemic immediately.
4. Fix local incidents without pausing the whole room.
5. Announce support pattern once so participants know what to expect.

Suggested line:

"If one table drops, keep discussing. We’ll recover that table locally."

Live monitoring rhythm

Use a fixed cadence instead of reactive monitoring:

• every 4-5 minutes: transcript freshness sweep
• every 10-15 minutes: summary coherence spot-check
• after each break: reconnect and silent-table scan

Cadence prevents blind spots and over-intervention.

Incident playbooks

Single-table dropout

1. local reconnect
2. one-sentence verification
3. swap to spare if unstable after one minute
4. verify single active device remains

Duplicate capture signs

1. check for two active devices at one table
2. retire extra device
3. verify clean stream for 30 seconds
4. recheck summary convergence next cycle

Multi-table latency

1. confirm broad scope
2. avoid random per-table changes
3. keep facilitation moving
4. escalate shared infrastructure path if persistent

Break management controls

Before break:

• ask tables not to close the active phone page
• remind one active phone per table

After break:

• run quick fresh-line sweep
• prioritize silent tables first
• verify capture before next task starts

Closeout reliability checks

• summaries are stable enough for reporting
• no unresolved duplicate summary artifacts
• major incidents logged with action and outcome
• retention/report ownership assigned

Post-session improvement loop

Within 24 hours, review:

• repeated failure modes
• average recovery time by incident type
• room/network zones that underperformed
• setup changes required for next run

Reliability improves fastest when each workshop feeds the next checklist version.

Reliability readiness score (optional but useful)

If you run workshops repeatedly, score readiness before each session:

• Mapping readiness (0-2)
• Device readiness (0-2)
• Network readiness (0-2)
• Response readiness (0-2)
• Closeout readiness (0-2)

A total below 8 is a warning that launch risk is high. This simple score helps teams discuss concrete weaknesses before participants arrive.

This guide is for...

Use this guide when you need consistent operations across repeated high-stakes workshops.

If you need deeper diagnosis of one failure type, jump to [Troubleshooting transcription latency](/guides/advanced/troubleshooting-transcription-latency) or [Device reconnection behavior](/guides/advanced/device-reconnection-behavior).

Related guides

• [Troubleshooting transcription latency](/guides/advanced/troubleshooting-transcription-latency)
• [Device reconnection behavior](/guides/advanced/device-reconnection-behavior)
• [Microphone replacement without data loss](/guides/advanced/microphone-replacement-without-data-loss)
• [Starting a RoomRadar session](/guides/setup/start-a-roomradar-session)
• [Run a multi-table workshop with RoomRadar](/guides/workflows/run-a-multi-table-workshop)