I have helped implement one project where we put in circa 60...

I have helped implement one project where we put in circa 60 surveillance cameras at a number of council bus hubs. Through that project I did hear that it is super rare (never?) to have anything caught live on the cameras - due to this reason the stakeholders agreed there was no need to operationally have anyone watch the cameras. My vendor on that project manages thousands of cameras for clients and noted to me more than once that people have this romantic notion that you will catch things happening real-time - but it just doesn't happen.

Probably due to this on this project, my client was mostly interested in asset protection, privacy of the public, and being able to react to reported issues (re. get the recording after the fact.)

The Icetana system must give a lot of false positives. I wonder how the alerting system works and how it works in with normal operational processes for whatever organisation is using the system.

The business case for commercial level surveillance systems as it stands is difficult to make - keep in mind you probably need to also arrange for fibre to the locations etc. which blow out costs a lot. When the benefits of a surveillance system are listed against what is a significant cost for implementation and support the business case to go ahead is often due to strong thoughts of a key stakeholder or two that it is required vs. how the business case should be measured against costs and benefits.

Due to the above, I do wonder what additional cost the Icetana services adds to what is (in most cases) a low value surveillance system for how much additional benefit gained.

I can some organisations being interested in smart surveillance - casinos for one - but I can't see the Icetana deep learning picking up features of surreptitious cheating activity in that environment - especially without a training set for the deep learning. Picking up features like fights etc would be easy, but it isn't exactly covering the business case for the casino to have the cameras.

I only just learned of this company - I am keen to learn more (especially where my assumptions are wrong) and I am probably just coming into this a little cynical due to the above mentioned project that I was involved with.

Regardless of what I noted on limited business benefits against cost - organisations still pay for surveillance systems (increasingly so) and a % of these organisations will have the romantic notion of catching things in real-time so I assume Icetana should have a business model that will work if they are not pricing themselves out of the market.