A.D.A.M – Adaptive Data Center and Migration Assistant
A Solution to a Real-World Problem
Description of the problem:
According to the Ponemon Institute Research report, in 2016, the average cost of a data center outage was $740,357. What’s more, is that research indicates that 93% of businesses who have suffered an outage for more than 10 days, filed for bankruptcy in a year within the event.
As we know, Data Center problems and outages can range from hardware, software, change, incident, network connections, configuration, service provider etc., and can cost millions to resolve.
Solution to help make troubleshooting easier:
ADAM - is a voice-controlled virtual assistant that can help troubleshoot your Data Center in normal day-to-day operations and during migration windows.
How?
Adam uses several AI techniques in the back-end. One technique is a neuro-symbolic approach to
troubleshot/diagnose issues within a customer’s environment. After
speech is converted into text,
NLP preprocessing techniques are used on the text, which then gets
converted into a deep neural network to predict a response. Finally,
the intent
prediction goes into an expert system. Embedding connectionist methods such as
neural networks into a symbolic-based planning systems allows Adam to not only see
the reason why a certain decision or output was selected or chosen,
which improves
explainability, but it also allows Adam to learn and reason about the world by
observing it. Here is just a preview. More to come!

Comments
Post a Comment