Can predictive analysis help MSPs and Technology Partners dramatically improve Uptime?

Everything we create as humans, buildings, art or technology - when things break or anything gets "out of place" we fix them. That worked until we invented aircraft. The airline industry changed the practice of “fix when broke" forever.

Today we all accept that everything needs maintenance. The world of IT and IP technology is not different. Even though our IT & IP solutions will not physically fall out to the sky and crash, they will certainly virtually fall out of “the cloud and crash". Humor aside - IT and IP technology are critical infrastructure and the nerve systems of our businesses and to be down even for minutes can be very harmful to the businesses affected.

The airline industry showed us that we needed to apply predictive analysis to all potential problems to anticipate the mechanical failures to preempt them and that measuring specific parameters leads to solid predictive analysis. For example, measurements of oil pressure and temperatures allows preventing engine disasters, especially if analyzing the rates of change and interactions between outside temperature, energy output and unexpected temperature rises.

At IPTechView^TM, we decided to take this same approach to IT and IP technology. We started collecting and analyzing critical system health data of all the equipment our customers use when building IP Technology solutions so that we could track every element of a project. The chain is only as strong as its weakest link. Initially, we gathered data that we could find on our own but we realized we needed a more complete picture and went back to our key vendors and asked them to proactively send us a standardized set of system health data using our custom createdIPTechView^TM API. This step led us to being able to unify all the parameters and have a common data stream of health parameters that we now can process for thresholds that we can generate using hard coded logics, and enables processing of this information by AI and statistical algorithms. The original vendors that embraced our API companies like DrayTek, Epygi, Fanvil, MOBOTIX and Overland-Tandberg with the data we were able to collect over the last two years enabled us to create a proof of concept and develop IPTechView^TM. The collaboration with our vendor partners and dozens of resellers and MSPs helping us with BETA testing over a period of almost two years were all instrumental in us taking our initial research from an experimental proof of concept to a large scale working system into a tangible universal solution for IP Technology in general with partners all over the world.

The end result is that today IPTechView^TM notifies partners of unusual conditions and provides warnings that will allow partners to take more preemptive steps to replace or maintain their equipment, change settings to ensure their customers systems will operate continuously at optimal levels, and try to eliminate downtime whenever possible. For us this is to truly “Break out of the Break Fix Cycle”. Our Resellers and MSP Partners benefit from being able to provide faster and better services and tighter SLAs and end-users are getting less disruptive incidents higher UP-Time and are overall much more satisfied with solutions managed by IPTechView^TM. IPTechView^TM notifies technology partners when system health parameters deviate from the thresholds and health zones that were established statistically over the past years or show behavioral patterns that show equipment is at risk. The threshold system we designed though is still as easy as seeing the temperature or oil gage on our cars and we built it so that each partner can fine-tune each device for each particular customer's application. A unit running in a hot, uncooled cable closet will have a different profile than a unit running in a cold server room. Like in our cars not every problem can and will be foreseen but having an active warning and alert system that notifies us and a systems visibility - the ability to see all gauges at once will enable partners to react much faster and help maintain a much higher overall up-time.