How we roll it out (onboarding)¶
Onboarding a center is a clear eight-step path from "we've agreed" to "it's working." There's no need to break anything in the teaching process — ControlAI fits into what is already there.
Step by step¶
1. Audit¶
We inspect the rooms and the center: audio quality from the cameras, NVR/DVR availability, recording retention period, internet speed, the clock on the NVR/DVR (details — in the chapter «What we need from the center»). The result is a list of what needs adjusting.
2. Add-on installation where needed¶
Where the audio is weak, we install an inexpensive external microphone; where needed, we adjust camera and access settings. These are small, targeted tweaks, not a re-equip.
3. Consents¶
We collect consents from teachers and, where required, from parents. The center presents For a teacher, ControlAI is a personal assistant: personal stats and a breakdown of their own lessons — and that is how it should be presented to the team. The consents explicitly cover the audio and video recording of lessons (details in the chapter "Privacy and trust").
4. Setting up the center in the system¶
We enter the center's structure into the admin panel: branches, rooms and cameras, courses and levels, groups, teachers and students, and — most importantly — the schedule. It is the schedule that switches on the automatic capture of recordings.
5. Configuring the methodology¶
Together with the academic manager (methodist), we upload the center's rubric (the observation form), the curriculum and textbooks, and set the review focus. Now "AI-nazoratchi" checks lessons against the standards of this particular center.
6. Trial run: one classroom¶
We launch one entire classroom — its first lessons on the schedule, from recording to a finished report — and look at the live result together. This removes any doubts and shows that everything works as it should.
7. Go-live¶
We connect the remaining classrooms. From this point on, ControlAI analyzes every lesson on the schedule and shows the results in the admin panel.
8. Ongoing support¶
After that, the center runs on its own, and we stay close by: we show the team how to use the panel, monitor processing health (through our service role), and help with tasks from the "Operations" section (camera offline, microphone needed, etc.).
What determines the timeline¶
Most of the time goes into the audit and the targeted hardware tweaks; entering the center into the system and configuring the methodology run in parallel — so we give the exact timeline after the audit; it depends on how much needs fixing. The trial run removes the risks before the full launch, so go-live goes smoothly.
Key takeaway from this chapter: rollout is audit → small tweaks → consents → entering the center and the schedule → configuring the methodology → a trial run of one classroom → connecting all rooms → ongoing support. The teaching process doesn't change; ControlAI fits into what is already working.
→ Next: 14. Privacy and trust — how data and consents are protected.