Wired Assurance: Day 2 – Marvis Conversational Interface
Marvis is a great listener.
With a conversational interface using natural language processing, Marvis is able to understand user intent and goals. Just describe your intent to the virtual assistant, and watch it get to work. See Marvis in action in this Wired Assurance Day 2 demo.
You’ll learn
How natural language processing helps Marvis understand your intent and goals
How Marvis can help you troubleshoot and resolve client issues
How to avoid having to remember specific dashboards or CLI commands
Who is this for?
Transcript
00:03 marvis is a new member of your it team
00:05 that you can actually work with
00:07 the latest enhancement is a
00:08 conversational interface where users can
00:11 see pending actions or inquire about
00:13 something else
00:14 users can click on suggested actions or
00:17 ask questions like what is happening
00:18 with this unhealthy access point
00:21 marvis is a purpose-built ai engine that
00:23 returns the relevant details about your
00:25 network based on log-mining
00:27 machine learning algorithms and an
00:29 overall rich data science toolbox
00:32 in this example marvis processes the
00:34 site's ap data
00:35 and concludes that an upgraded firmware
00:37 image can make the ap healthier
00:40 the user tells marvis yes and that's it
00:42 no need to search for the firmware or
00:44 upgrade button
00:46 marvis enables ai-driven support to
00:48 proactively resolve network issues
00:50 with live feedback and real-time data
00:53 here marvis verifies that the firmware
00:55 upgrade was properly executed
01:05 marvis's nlp has been enhanced with nlu
01:08 so it can understand what your intent is
01:10 even with misspellings
01:12 ask marvis to troubleshoot a specific
01:14 client by selecting it from the list
01:16 which yields the client's issues and
01:19 actions the user can take
01:21 this client appears to have some dhcp
01:23 issues
01:24 so asking about the failure timeline
01:26 yields a better understanding of the
01:28 history of the problem and when it
01:29 occurred
01:31 if you would like to see more marvis
01:33 pulls together supporting data for the
01:34 problem
01:35 and allows the user to dig in deeper
01:38 marvis contextualizes your intent to
01:40 show you what matters
01:41 like the exact client experiencing
01:43 issues and every time it failed to
01:45 connect
01:46 if the problem cannot be resolved marvis
01:48 can automatically
01:49 open support tickets in other systems to
01:52 get more eyes on the issue
01:54 the key takeaway of marvis is to focus
01:56 on making user experiences
01:58 better with streamlined operations to
02:00 quickly resolve problems
02:01 before users know they exist
02:20 you
