Importance of Time Series Insights

is going to present these two very interesting topics to the audience

And I believe you guys will have great time over the next two hours

And Deepak is a Microsoft MVP in Azure, Microsoft Azure in Canada

And also he is an MVP in C Sharp Corner. So he's writing lots of articles and very busy man attending all the sessions and conferences

We're really thankful to have him today in our webinar. And we are also fortunate to learn lots of things from Dibak because he is a 10 to 15 years experienced IT professional

And he has worked in all over the world, like in the U.S., lots of big companies in U.S., Canada

So it's really fun to know the experience of this kind of person who has that much experience

who has overcome all those obstacles in his career and now is successful MVP and also works with Canadian government

So hello, Deepak. It's your time to shine now. Please educate us with your knowledge

And we are really interested to learn about the importance of time series insights

So, yeah. Thank you. Thanks a lot, Rahat, for your nice word

Hi, everyone. I'm Deepak as Rahat Tal and I'm Microsoft Azure MVP

C-Sup Corner Chapter Leader at Regina. Regina is pro capital of province of Saskatchewan so she's up Connor MVP also

and Regina chapter late where we are we are hosting our events tech events very

oftenly so I blog at deeper cosmic calm if you have any questions or you want to

contact or for example you you want to listen this recording again so deeper

cause it calm you could go to that website and have this recording so I

tweeted thing for the book that is my Twitter handler I'm most happy to

contact and connect with you guys with any questions and any information you

have we have our YouTube channel Deepakosik Microsoft MVP so hold this recording just FYI

this recording and these presentation would be available at our website as well as a subconer

the video will be available and YouTube thanks a lot for joining today's session I'm we I'm happy

and we are fortunate to have Rahat he's data scientist he was working earlier

with University of Saskatchewan and now he is AI head at Montreal Canada we both

are based on Canada so if you have any questions floating around that or you

could connect with Rahat or me any point of time so today my session is where I

would love to talk about is a zero time series inside and that's very very close

to my heart there are a couple of reasons why we are moving or there is

trend towards as your AI or you could say time series but before that I will

touch base couple of questions and basic fundamental then we will do a demo and

I will take question and answer and then we will summarize so first of first and foremost thing which we always thought about what the heck is

IOT my kid who is 10 year old he came to one day dad what the heck is IOT why a couple of guys

You know talking about higher T and looking over different things so I I I talked to him okay IOT is when one physical device talk to Internet is

called IOT so for example your band health band which talk to internet and

say okay your heartbeat is heartbeat is increase or you have the sugar level

increase then probably that data will go to internet and there will be some actionable insight

and probably you will get an email or sms okay now you have to take your sugar medicines or

aspirin so that the thing uh is about iot where physical device talk about internet so nowadays we are thinking iot is new but it it uh iot was

technically invented or put forward by kevin in 1999 we could we could see it's a quite few years

uh from now but uh yeah last decade or if not decade to be very precise from last five years

ai and time series they are playing a very good role in our day-to-day life

just taking this example look over this picture and what you think about that

for example if any morning tomorrow you get up you go to wet wing machine and it will say okay I

will I will share the detail of your weight at internet I will tweet about that or I will email

or I will post it on Facebook you are hundred pounds or you are hundred K so

how how that would work so that's the IOT because your physical device for

example your weight-fitting machine know about you a so it could see B it have

data it has weight and see if it has connectivity then it's a IOT device

earlier we were talking about a lot of power supernatural power I grew up with superheroes

Ironman etc but nowadays it's about IOT so if you have knowledge of IOT so probably you have no limit

of sky you you will sign down the time so this is kind of suggestion from my

side if you want to explore your expertise towards IOT that is the good

area where you could look over so just take a real-time example how now

eventually I will be more detail and with more example I will share one case

study working case study live case study and one Canadian client case study where

I propose a solution to them so I will address that eventually now I'm

switching my gears to a small IOT time series for example in New York if you

had been a New York and you hire a cab you have opportunity to go to this URL

and look over time series data if you look over time series data you came to

know at what time which taxi has started which is the passenger at which direction

it went how much it cost and all that detail so by owing to these if you go to

their website you will came to know okay if you want to share a ride you could do

that because next taxi is coming I want to show that my browser so so you have

that opportunity you will look over that detail so if you look for my screen yeah

so I could see so this is the URL of New York website New York taxi where you

have the historical data and you could do actionable performance over that so

that is one kind of real-time scenario which is which has been already

implemented in our day-to-day life for example let's what if we have data how

how could we perform action so on any data we could figure out the anomalies

and on the basis of anomalies we could perform certain action so what is the

anomaly and only is the problem of detecting anomalous pattern for example

our heartbeat or sugar level if that increase or decrease that is one example

let me give another example let me give some real-time example of one of the

power client in Canada and as that is trans Atlanta if you look over my screen

right now this trance altar is the Canadian power client they have

implemented as your time series with Azure event hub so that every time

wherever there is break in any power grid or power fan I will be not using more technical terms related to power client to make you understand but I giving you the solution of the of how IOT is implemented so for example

of fan is producing some heat and some vibration and if we record that data we

came to know okay this fan is not working or this fan has some anomalies if I'm

saying anomalies so probably after a couple of months it's going to stop

working in that case we have that actionable item where we could figure

out okay this is the problem this is going to be happen a problem and we

could take action so it's not it's definitely not a solution after a

problem it's kind of we are doing actions before a problem so that's

another example which I have given if you want to have this detail you could

have this detail from stay tuned with our website you have this detail or

there but I would love to share with you also this URL where you could find

Canadian client example Canadian client example I'm putting that on chat window

I won't I would like to make it very interactive and I would like you

understand maximum so you read about this and you will came to know how time

series in sight would be helpful for you guys so we talked about transit Atlanta and we have the anomalies where we could

figure out okay these are the anomalies we can find out now more zero down to as your time series

and look over the as your time series is or as your time series in site is as your time series

is manage ytics if I'm saying ytics you could also use power bi but you could also leverage

your selfie the channel its ytics visual lizes and service that is simple

to explore for example nowadays we have thousands and thousands and thousands

of gadgets sending huge amount of data to the cloud every minute and second and

it's almost difficult for us or any person to yze on their data so the

azure times series inside its solution of that problem and we could do that now

it's time for demo as I promised we will do couple of demos it would not be one

way so before demo just setting again this is the website where you will find

the recording and you will also find on C sub corner as well as YouTube channel

so let's go it's time for demo and I'm writing time series so this is my azure

portal I go to azure portal and write time series so it's give me option time

series inside I said oh yeah that's time series inside huh I would want to do a

demo so I'm doing here demo webinar webinar demo I use my sponsorship

subscription I'm using existing resource group but no Damien one I would say demo

not demo for summit where which location I would like to use it's always

recommended to use the location nearest to your geographical location you could

look for price detail how much it will cost by default I'm using standard one

which cost me 182 dollars per month how much capacity we want so it's you could

see by default in Canadian dollars the value is changing but I'm using a

minimum 1 million events per day for me that's the event I'm looking for as soon

as I create it it will start building and deploying my deploying my event and

But before that I have another my demo another time series demo over here where I have already did that so that we could save our time and after deployment it would come like this and this beautiful screen of time series

insight will come up if we look over gracefully it's what interval time we

are looking for which event we want to address we could give the more details

if you want to give some okay less than 500 degree centigrade something like that for heat etc so after

that once you will have this demo ready your a platform ready you will click and you could

yze your screen okay these are the events which are the events I would like to look over

and after that you could yze you could do heat map which is my personally

favorite heat map over here and you could look over the details in table so

after you have all these events and charts ready you could perform your

ytical parts or give share this data to business hey this is what we are

facing problem or this grid is going to be break and that would be a another

solution I would not like to share the detail of client here but I would like

share screen how this is implemented this solution has been implemented

suggested to a client in Canada one of our client is facing problem where they

there's power fans are most oftenly stop working and that cost them thousands or

billions of dollars in addition to that they have to pay some penalty and they

have to buy some energy from different source so in that case I said okay save

your vibrational data from sensor to azure IOT hub one of the cases for IOT

hub is to gather telemetry from the device so if you look our next screen

you will find what the IOT hub do IOT hub is the hub where we have million of

connected device we can do device to cloud messaging and cloud to device

messaging for that client my solution was okay from this fan just connect with

IOT hub and IOT hub will pass the data to the cloud and we on portal we could

yze that so we could simulate wind turbine which emits the data we can

provision some device if there is any telemetry telemetry is the problem event

processor receive telemetry then we you know forward that detail for benefit so we get nearly

real-time insight we start in seconds and global view of IOT so that's the benefit and as we talk

about this solution has been implemented by couple of clients couple of clients if you look over IOT

stream ytics if the URL which I have shared just now a couple of clients a

couple of cities smart cities like Barcelona they had implemented the time

series inside inside for their traffic system for example if I'm at Barcelona

and I miss one bus it's not I'm looking or some static information at real time

I look over where is the bus how much time it will take and in that time can I grab a coffee if bus is

coming after 30 minutes I could go and grab a coffee in debt or make a call or send some emails

that kind of details so with that I would like to say thanks you guys for joining today this

session now it's over to Rahat who will talk about cognitive services and as I

promised you will find these detail and case studies at different channels which

is my website my blog YouTube etc so thank you over to you Rahat thank you

Deepak for your nice and very informative session. I had, like, I know about IoT, but I will