Massive centralization & decentralization
We are proud that our
product does both and that exactly is the need of the day.
When product offers Overall Equipment Efficiency at
fingertips, the sensor data need to reach cloud servers for real-time analytics &
ML. Same is the story when third level backup is offered in the cloud.
Now when it comes to Vision Neural Nets, demanding tons
of video frames reaching cloud servers, it is just wrong positioning. Doing this
would invite all sorts of different challenges including band-width challenges or
sometimes, just becoming a headache to SDWAN. So, we do de-centralization here by
making the AI work at edge, that too in $80 hardware.
By the way, does that mean, after deployment AI model will get outdated? Answer is NO. Our Neural Net Algorithm could mark outliers & trigger re-training and updated Neural Net reaches edge. Thus, updated model reaches edge automatically.
This competence of our algorithm, in turn fosters our relationships with customers that lasts for a long time.