Up next in 10
Join Henk & Amy in this 30-minute talk show spotlighting people from the community that work with (Microsoft) AI technology in real life. We talk about their job, their view on AI. Focusing on the person in the AI field, rather than focusing on the tech. The Tech is the tool, and the person is the hero of the story. With this show we will get insights how people are using our tech, what they are building and open an opportunity to get real life feedback.
A bit of AI Show Page: https://www.c-sharpcorner.com/live/a-bit-of-ai
C# Live - Dev Streaming Destination: https://www.c-sharpcorner.com/live
C# Corner - Community of Software and Data Developers
https://www.c-sharpcorner.com
#abitofai #talkshow #podcast #csharpcornerlive #csharpcorner
Show More Show Less View Video Transcript
0:30
Thank you
0:59
Thank you
1:29
Thank you
1:59
Thank you
2:29
Thank you
2:59
Thank you
3:29
Thank you
4:30
Thank you
4:59
Thank you
6:00
Thank you
6:29
Thank you
7:55
Thank you
8:24
Thank you
8:54
Thank you
9:24
Thank you
9:54
Thank you
10:24
Thank you
10:54
Thank you
11:24
Thank you
11:54
Welcome to the A Bit of AI Show with your hosts, Hank and Amy
12:25
Hello everyone and welcome to the first bit of AI show with myself, Hank Buhlmann and Amy Boyd
12:36
We both work at Microsoft and have a passion for AI. And in our journey and meetings with many people
12:44
around the globe, we realized that there are so many different roles within AI solutions
12:49
from pre-sales to the people that are creating the models. So in this show, we invite people from all over the world
13:01
who are professionals in the AI space to have a chat about what they actually do during the week
13:10
This is the first episode, and we couldn't be more excited after a long time of planning
13:15
for this actually to take off so that we can talk to our guests
13:21
Because this is the first episode, let's talk a little bit about the format of the show
13:26
We have four segments in the show. We will start with highlighting some nice upcoming events
13:33
Then we go to our main guest to talk about her or his life in AI
13:38
We close the show with our second guest to talk about an open source project
13:44
and how you can contribute to that. And finally, we dive into an AI learning challenge
13:51
that we will do weekly, which you can all find on our website, abitofai.show
13:58
So let's start with some upcoming events. I want to highlight in this episode two events
14:09
In April, the Global AI Student Conference is taking place. And this is an event created by students, but not only for students
14:19
Everyone can, of course, join and learn. And there is still time, I think, four days to submit some sessions
14:27
So more information can be found on AIConf.education. And of course, everything we talk about today can be found on our website, abitofai.show
14:37
Another great weekly event I want to focus on is AI Talks created by Sami
14:46
He's an AI MVP and every two weeks he invites guests to his one-hour show to talk about a
14:53
cool AI project This you can find on globalai So I think it is time to welcome our first guests
15:07
So let's get started. Hi, Terry. Hey. Hi, Amy. Hi, Hank. Hey, Terry
15:16
Thank you so much for joining us. Thank you for having me
15:19
It's very exciting. Glad to be here for the very first one. so no pressure yeah no pressure and who knows where this could go right in the
15:30
end we might be famous on the internet and cool well let's get started because
15:37
we don't have that long but we want to make sure we completely dive into what
15:41
you do and who you are Terry so let's start there tell us a little bit about
15:45
yourself. Cool. So hi, I'm Terry. So I run a AI consultancy called Advancing ytics
15:55
I'm also a Microsoft MVP for AI, which basically means that I probably talk at too many events and
16:01
probably don't do enough real work. And yeah, I live in Devon in the UK with my wife and my
16:08
two kids, which has been really interesting during the whole lockdown period
16:15
fabulous oh no that's uh devon though it so i'm from the uk same as terry um devon's a beautiful
16:23
place that's uh what a nice place to to be situated especially at the moment and also
16:29
coming to the summer um hopefully others will be able to get down there at some point um so let's
16:35
uh yeah let's let's kick off um first things first i always think this is the hardest question to
16:42
answer but also in the most insightful question for people when it comes to roles um if you do
16:48
shadowing and stuff like that you'll tend to kind of feel this but um there's so many different jobs
16:54
and that's kind of in AI and that's why we've kind of started this show uh so can you tell us
16:59
what does your day look like an average day so um you know how is it you start what do you what
17:07
you end up working on all that kind of stuff yeah i think if i if i start at the at the right at the
17:13
beginning it's a manic morning uh getting children ready to ready for school and other things um and
17:20
then when i start my day job um so things have changed being remote but so advancing ytics
17:27
we're a consultancy and we help customers solve data engineering problems and ai problems so an
17:33
average day can be so different depending on which customer we're working with and which project
17:39
we're working with and it could be that one day is spent on a project actually with a customer
17:44
helping them solve a particular problem and that might be looking at you know how does a customer
17:50
churn when's the customer churn and leave their platform it might be looking at how we apply
17:54
hyper personalization to their application so that they can better understand what the next best
18:00
action is for somebody to take on their platform. Or it could be, you know, looking at something a
18:05
bit more advanced, like how you do computer vision, looking at, you know, warehouse imagery
18:10
a whole load of different bits and pieces. But then when I'm not kind of doing the technical
18:15
side of things as well, you know, we spend a lot of time around upskilling the team
18:20
as well as doing, you know, internal hacks and trying to get, you know, trying to keep up to
18:26
date with a lot of a lot of you know pretty much what's happening in the field so it can be
18:31
completely different every single day most of the time it's with customers collaboratively
18:37
helping them build either um either a deep learning application or even just a shallow
18:43
learning application depending on you know what it is that they really need so varied
18:48
i was gonna say that's the i think we kind of lost you amy oh no not now you're not in the first episode
19:07
come on laptop stay with me stay with me thank you terry for sharing the the variety the variety
19:16
in my life is whether uh my laptop on my internet work that's always fun um so the the variety is
19:23
the key though right do you thrive on variety do you think you would you be as excited if it was
19:29
the same every day no absolutely not i'm one of those people that after three months on any project
19:35
is i'm like crawling up against a wall i really try and like i need something else uh so no so i
19:40
think variety is is the key and with that you know we work on so many different projects
19:45
in different um you know different industries and different verticals it could be you know helping
19:51
a startup build that personalization model it could be helping an insurance customer understand
19:57
how they look at claims fraud there's so many different avenues and because it's so varied and
20:04
you know projects are typically somewhere between you know three and six months you're just
20:08
constantly changing and that's what keeps it fresh and exciting and also when you try to when I think
20:14
about the skills that you need by constantly changing it means that I have to keep my skills
20:19
as sharp as possible and know you know enough about a variety of different approaches which is
20:25
always fun and always challenging but you know that's that's how it goes
20:30
well i know um you think that first question on average work day was hard so henk has the best
20:38
question that he ever put on one of our scripts i think take it away i can't i can't say this one for
20:43
you oh yeah yeah yeah so my question would be what is the most annoying thing about your role
20:50
in the AI field? Oh, that's such a good question. The most annoying, I think there are lots of
20:59
things that annoy me, but I would say one of the main things that's annoyed me for years
21:04
is the kind of misconception that you need to have a PhD to get into machine learning and AI
21:11
And I think it's such a huge fallacy that you need a PhD or that even having a PhD would make
21:18
you better at understanding machine learning problems and i think for some it does there are
21:24
some people i know who have done phds in machine learning in reinforcement learning and they are
21:29
phenomenal oh oh no camera can you still hit me my camera decided to switch off temporarily
21:36
so we can hear you okay there we go back um but yeah i think we work with so many different teams
21:44
with different skill levels and yeah we work for so many skill different teams and different skill
21:51
levels and it makes for a completely different you know you need a diverse audience you need a
21:58
diverse team and I think having people with deep technical backgrounds is one thing but then you
22:05
know there's much more to it I'm changing camera so that you can get me back I was gonna say for our
22:13
video stream, this is perfect, but for our podcast, this will not affect you, there we go
22:19
Right, one camera's down, backup camera, here we go. Very good on having a backup camera super That what we like to see So yeah I think the misconception that you need a PhD is one thing that really annoys me You don need you don need a PhD
22:37
You know, there are so many people that come from different backgrounds and it's super important to have different backgrounds in a team
22:44
And, you know, there's there's always been the concept of people who are unicorns who can kind of do everything
22:51
and it's nice to have those people but having a multifaceted team is so much better
22:58
than having one single unicorn who is probably really stressed out and trying to do absolutely everything
23:04
So I think that's one of the things that really annoys me. You handled that question well, if I'm honest
23:15
Yeah, that was good. So I'm also wondering, what is the last skill you have learned
23:23
Oh, the last skill I have learned. So I would say this has just been a continual process
23:34
but the last skill, and it probably is a set of skills or an entire discipline on its own
23:40
but we spent a lot of time really focusing around MLOps and machine learning, you know
23:46
how you apply data ops and DevOps to machine learning. And, you know, three, four years ago, we were really trying to define patterns around how that would work
23:55
And all the tools are getting more and more up to speed. So I think constantly what I'm doing is taking things that we built out manually by applying software development techniques to machine learning
24:07
And now looking at how that arises in new tools and actually figuring out, OK, well, how do we augment what we used to do with these new skills so that we can ultimately productionize machine learning models faster and more consistently
24:21
So that is always a continual area of exploration for me, because, again, it's so multifaceted between how do you deploy a batch model compared to a real-time model that's sat behind, you know, a million requests a minute
24:36
It's, yeah, so different and huge, huge variety in there. Yeah. Gosh, that's so..
24:44
One of those ones, right? Software developers can really get into this space
24:51
because ML ops, because security, because, I mean, at the end of the day
24:56
it's still tech, right? It's just a different kind of form of an executable
25:01
that's made in some senses. So I'm so, so happy that you kind of pressed on that point
25:07
Now, Terry, me and you have done this before. It is our quickfire round of questions
25:14
Now, what I love about a quickfire round is it should be the first thing that comes to mind
25:20
So it's the most natural thing. When we ask for something, try to keep it to about a sentence
25:26
and then hopefully the variety that we get from all our guests will show the extreme variety that
25:32
we have within machine learning. Are you ready Terry? I am ready, yeah. Good stuff, let me just
25:39
shuffle in my chair, right I'm ready too. So our first quick fire question is what was your first
25:45
programming language? My first programming language was SQL, maybe PHP but I don't tell people that
25:54
one. But SQL definitely was the first one. What program language was used in the last project you worked on
26:03
So we work an awful lot with PySpark. Pretty much everything we do is in a distributed way
26:09
So PySpark. That's an interesting one. One to note down. Okay, next question. No more than a sentence
26:17
if possible, Terry. What was the last thing you learned in AI? Graph, neural nets
26:28
Good one. Favorite event in the AI calendar? Data and AI summit
26:40
Oh, nice. Good choice. And then final question. This is nearly over
26:48
What area of AI is on your list to skill up on next
26:52
I am really interested in Transformers and how, you know, there's so much great stuff at the moment around Transformers and adversarial nets and using Transformers to solve, you know, so many problems
27:06
NLP, audio imagery. Yeah, Transformers are definitely an area where I'm looking
27:13
Ah, interesting. So that is our quickfire round done. Everyone can breathe
27:18
and hopefully that gave some people some insights into kind of just those immediate responses love
27:25
that you said SQL at the start right so you've come from a like a pure data background have you
27:31
is that where you've kind of started before machine learning yeah yeah absolutely a long
27:37
a long time ago pretty much starting off you know doing doing a lot of kind of data ytics
27:42
and then slowly augmenting those skills and you know having to true up in all other areas
27:48
and really kind of focus on that, which, again, I don't have a PhD. I mean, my background is in media studies
27:54
You know, that was my undergraduate. Yeah, I've since done Masters and stuff in machine learning
28:00
to further increase what I know. But, again, you just don't need that kind of stuff
28:05
If you're interested, by all means do it. But, you know, you don't need a software engineering background either
28:11
Again, I think different people, different skills really, you know, help make a great team
28:18
Yeah, and we've got just over a minute left. Can you believe 15 minutes has gone already? This is
28:24
where we have to jump straight into those questions with you. What's your learning process
28:31
Recently, I've found picking up new technologies, Hank, we were talking about this actually
28:36
recently, that it's like, whatever you've worked on recently, has kind of probably been maybe
28:42
something you knew kind of baseline theory about, but realistically, you always have to research and
28:46
try new things so do you have a a learning style if you're going to take on a new technology
28:52
yeah I mean I think yeah definitely I mean to do any of this you need to be a deep learner and you
28:59
need to be able to do that kind of deep work for me it's always I read as much as I can I watch an
29:06
awful lot so I watch quite a lot of you know videos about papers or read papers and find
29:12
interesting things there but also I just quite love reading technical books and then just trying
29:18
stuff out I'm a I'm a complete doer have to learn through doing and so trying things out repetition
29:23
is really you know how I get stuck into things I'm so glad we ended this section on a deep
29:31
learning pun can we get those sound effects for next time Hank that'll be
29:36
you can do a little one Who doesn't love a false clap
29:45
Thank you so much, Terry. We've had our 15 minutes with you where you talk through very gracefully your role
29:55
the variety in there, the ability to work in diverse teams and the
29:59
to always learn. Any last tips or tricks you want to share with our audience
30:07
I would say just, you know, find something that you're passionate about. If you're new to this
30:12
area, focus on machine learning on something that you are passionate about and that you know
30:17
and you'll always understand things so much better. Love that. That was perfect. Well
30:23
thank you so much, Terry. We'll say goodbye for you for now, but we will see you again in a little
30:28
bit and we will move on to our next guest. Hank, it's over to you to put our next guest in the hot
30:34
seat, I think. Yes, and maybe this is a good time to talk a little bit about the AI, the bit of AI
30:45
cafe in between, because there are a lot of questions in the chat. We have like 61 live viewers
30:52
that are asking questions. But due to the 15 minutes show we have, it is kind of hard to answer
30:58
that questions in real time so that is why we have the um a bit of ai cafe where we can hang out
31:05
after the after the show so for me that will be at 10 30 in like 10 minutes the link is on top of
31:14
the website and you will join the teams meeting and there you will find terry and arafat and me
31:20
and amy and our reactor team and we can just have a casual conversation and that is where we will
31:26
answered all those questions you've been asking like what's a good entry job would do
31:35
what do you see is the next level of ai applications to come forward post pandemic
31:40
so those questions we will answer all there and i think um maybe terry can answer some in the chat
31:46
already and then we can have a conversation in our cafe you can see it as the the pizza after the
31:54
after a meetup kind of virtual pizza depending on your time of day i don't think i want to eat
32:01
pizza at 9 20 in the morning but it's happened pizza isn't any time snack it's good it's an
32:09
anytime snack cool so but we are on time um so terry thank you so much for joining
32:17
and let's switch to switch to Arafat hello hello good evening the internet demons have found us we can hear you
32:40
oh can you hear me now yes can you hear me now all right yeah perfect there we go you're back
32:53
you're back don't worry we we just roll through these things roll through them at this point a
32:58
year in. Okay so let get started straight away So very very briefly because we only have five minutes
33:13
with you. Who are you? And what do you do? Okay, yeah, good question. So I am Arafat Tehseen. I am
33:22
a solution architect at EY and a Microsoft MVP for EY. So I do a lot of things with bots
33:30
applied AI, Microsoft applied AI, and the platform as well. So if you want more information about me
33:39
you can just go github.com slash Arafat Tassin, and you can find a lot of information about me, yeah
33:49
Nice. So can you tell us a bit about the Open Source Project? What does it help people achieve
33:54
yeah so i'm i normally look after two of the projects um one of the projects which is a
34:03
cognitive rocket so it is on my own personal repo where my own stuff such as
34:09
or or the talks or the presentation or any blog post which i'm writing so i put my code over there
34:17
broader audience and if they want to get any benefit out of it um the other one which i'm
34:22
going to talk about today is the bot builder community project. This is a central repo for
34:28
the community resource. It has got all the community resources for the developers who are
34:33
creating bots using Microsoft bot framework. So we have got a bunch of stuff like extends
34:40
recognizers, customer adapt, like you can surface your on Google Assistant, Zoom, Ring
34:46
central and and others yeah fantastic there's um there's a lot going on in that repository i've
34:54
worked with that repository so this is exciting to have you on the show um what type of contributions
34:59
are you looking for is there anything specific or or what's already in there that people have
35:04
contributed yeah so um who have contributed it's it's a very big so i'll go a little bit
35:14
history for that. So in 2016, when I personally started with Bot Framework and found out that
35:22
there are plenty of resources which are scattered here and there, and then I, along with
35:28
some other Microsoft MVPs, who is a part of Microsoft now, Michael, Stephen
35:36
we came together, we came to the point that, yes, we should have a single repository
35:40
for Bot Framework community resources. So we found one and we've seen there was a lot of
35:49
there was a need for this and we got a lot of support from Microsoft. We now have more than
35:55
50 contributors on this project. So you can see a lot of things are
36:01
happening over there. In terms of taking part how can you take part into that repository or you can contribute to that project First you need to be a developer or have a bot framework knowledge
36:18
and the passion to share with the community. If you're not a developer, then don't worry about it
36:25
You can see a lot of developer resources. If you're not a developer, if you are a no-code
36:32
or low-code expert and you are enthusiastic about bots, you can always come to me, talk to me
36:37
and we have a bunch of stuff for you still in our Bot Builder repo
36:43
which needs to be looked after. So the URL for the Bot Builder community project
36:50
is github.com slash botbuilder.com. And whenever you will go over there, you'll find a couple of people
36:57
So you can contact either me, Gary, or there are many people over there listed
37:02
so grab anyone and start talking to them how you can contribute to that yes fantastic
37:12
and so yeah all of those links are available on a bit of AI
37:16
dot show so if you head there you can find all the great
37:20
information that both of our guests have been sharing Hank is it time for us to wrap up
37:29
are we on time You have exactly three and a half minutes to talk about our last segment, about our learning challenge of the week
37:43
Yes, I was going to say, so as we were prepping, because obviously we do chat with our guests before they come on the show, and we actually know them in this case
37:52
So it was really, really easy. We have put together a bit of a challenge for you
37:57
So if you go to aka.ms slash a bit of AI dash learn, we have created a collection and a cloud skills challenge with two different modules
38:08
One of them relates to Terry and it's all around ML Ops
38:12
So there's this amazing sort of getting started module. So if you're new to the area, start the machine learning lifecycle with ML Ops
38:20
And then we also, I added in a bonus one because I felt like it fit really well with the open source repository from Arafat as well
38:29
And so we added in create a bot with bot framework composer
38:33
Also, Gary, our program manager here for Composer at Microsoft, will be incredibly happy that we put that one in there
38:42
So these are all in a collection. So there's just two modules and you can go and do the cloud skills challenge and take part
38:50
and just learn a little bit more about what both of our guests are working in
38:55
And so, Hank, can you believe it's the end of our first episode already
39:00
I don't even know where 25 minutes of my life go anymore when these things are so quick
39:05
It went fast. It did. But a lot of information. A lot of information
39:13
Yeah and obviously quick reminder if you want to join us in the cafe uh you can go to aka slash a bit of ai dash cafe we called it a cafe because it it going to be chatty it going to
39:27
be um open and respectful and uh you can meet our amazing guests and you can have a chat with us as
39:34
well if you if you want to um but on that note uh i just want to say thank you so much for joining
39:41
us today i hope you've enjoyed the show and keep in touch we are on twitter as well um if you have
39:46
any feedback if there's something you want to see on the show uh do let us know because it helps us
39:52
start scouting out our guests uh who can give you that great information uh we're here every thursday
39:58
um so it's a 9 a.m gmt uh i'm based in the uk it's 10 a.m ct it's a 2 30 ist and it's 8 p.m
40:07
AEDT so we cover a nice section of Europe and Asia. Come and join us in that bit of AI cafe
40:15
If you want to re-watch this episode go to a bitofai.show everything is there for you
40:20
and so I just want to thank you for watching a bit of AI show with Henk and Amy
40:26
Thanks everyone and see you next week. Thank you
41:07
Thank you
41:37
Thank you
42:07
Thank you
#Energy & Utilities
#Computers & Electronics
#Social Issues & Advocacy
#Green Living & Environmental Issues
#Ecology & Environment
#Climate Change & Global Warming
#Waste Management
#Green & Eco-Friendly Shopping