Business Applications | Technical keynote with James Philips

Business Applications | Technical keynote with James Philips


-We just can’t lose to
our competition just because
we have empty shelves. That’s where Power Apps
comes into play. That’s where Flow comes into play. That’s where Power BI
comes into play. They’re helping us
ensure proper execution. They’re helping us
improve on execution. They’re exposing
where we’re not executing. -The Power Platform allows us
to try so many things at such a higher rate
than we ever could before. Take these best practices
and the applications that really have legs
and get great user adoption, and specifically create
big business value, and scale them globally. -My hope with more educators
building their own apps is that we’re able to spend
more of our time in front of students teaching and more students
being able to learn because everything is either
efficient or automatic or we’re creating
or the students are creating and that’s really that
highest level of learning. -Please welcome Corporate Vice
President Business Applications Group, James Phillips. -Good Morning. Thank you.
Thank you. Thank you. Sorry for those who are filing in. We ran a little bit late,
but we’re going to keep the trains running on time here this morning. We’re going to spend about 75 minutes
talking about the Power Platform. It’s stunning how much growth
we’ve seen here. Hundreds and hundreds of percent growth in Power Apps, in today’s new announcement
with Power Automate, with Flow. Just amazing sort of
community development and an embrace
of this platform. Thousands and thousands
and thousands of applications being created
even at single companies. It’s stunning. We’re going to talk about why today
we’ve seen so much growth. We’re going to talk to
some of our customers. We’re going to go through
a bunch of demos and show off the capabilities, but I just wanted to put
these numbers in front of you. There’s massive momentum here and it’s across companies
that span every industry and the globe. This is happening in small
organizations and large organizations across retail,
manufacturing, transportation. It really is allowing just
about every company in every industry to embrace
this citizen development movement and to help digitally
transform their organizations. Sayed talked this morning about all of our technologies
kind of starting with Azure and working his way
up to Dynamics 365. I’m going to spend a little bit
of time on that as well before we get into
the Power Platform because I think
it’s really important that we sort of recognize
how all these pieces fit together. This is a unified stack of capabilities, and we’ve been
very intentional about doing
this integration work so that you can apply
the right tools at the right time and use them all together
to great effect. I want to tell you sort of why
we’re doing what we’re doing, the vision that we had
almost a decade ago now. It’s amazing how time
has sort of flown by. But if you sort of step
back about 10 years when we started on this journey
to the cloud at Microsoft, we had a singular vision,
and I’ll get to that in a second, but it really ultimately sort of
culminates at the application layer. I mean, all of the technology
that we have eventually works its way up
to applications that people use. If you think about the applications that we’ve been using in business
for decades and decades, I’m going to argue
they’re no different, right? They haven’t really changed
fundamentally. This is a picture of a TWA
call center in 1970. If you were to walk and look
at that screen and then take someone
and move them to 1990, move them to 2010, they would see more difference
in hairstyles and clothing than they would
on the screens. It’s not a joke.
I mean, it’s forms over data. Sure, in 1970 it’s monochrome
green or amber, in ’90 it’s a fat client with a Sun system
run in Siebel probably on the back end, in 2010 it’s in a browser, SaaS delivered,
but the same app literally. Nothing has really fundamentally
changed in these applications. If the business applications
are what sort of embody your business processes,
if they haven’t changed, then your business processes have been
fundamentally unchanged as well. Very reactive. You think about perhaps
you’re a manufacturer. You sell capital equipment
to someone who installs maybe
a milling machine in a factory. That milling machine fails. Now you’ve got a crisis. Someone picks up the phone.
They call a service center. Someone picks up the phone. Now an application opens,
now data starts getting created, and eventually you can react
and roll someone to go bring
the production line back up. Super reactive,
but that’s all we could do. There really wasn’t another way, but there’s something
that’s fundamentally changing and this underlies
everything that we’ve done at Microsoft, everything,
across every product and service. That fundamental change
is that data is starting to come out of everything
and it’s coming first. It’s coming out of every camera. It’s coming out of every
retail environment. It’s coming out of every factory. It’s coming out of products
and services. All of this data is flowing, not because someone opened an app
and typed something into a form, but because things are increasingly
instrumented, telemetry. All of this information is flowing
and that ultimately allows businesses to transform
from reactive to proactive. You think about the case
that we just talked about. Now you’ve got that milling
machine in the factory sending a continuous stream
of information about how it’s being used, the health of the device
or the milling machine. That lands in an anomaly
detection model in Azure, can predict that there’s
an impending failure. Now someone can go do
a maintenance call before the thing goes down
and keep the customer happy. It really flips entirely
the business process. It takes it from you have to wait
until something goes wrong to predicting what’s going to happen
and getting in front of it. Completely different. For the first time, applications,
sort of said this morning, are becoming this data first driven,
data intelligence action. This is happening across
the entirety of the business. Satya talked about
the digital feedback loop, customers, products,
people, operations or assets. All of these four corners of your
business are sending data continuously. Every time someone updates
their LinkedIn profile, visits your website, that piece
of equipment in the factory, on the product site,
all of this streaming information, that if you can harness it,
harmonize it and reason over it, you can really understand what’s happening
with your products and with your customer. You can then take that understanding,
for example, of your customer. We demoed customer insights
this morning. That 360 degree view of the customer
can drive intelligent engagement, which of course creates even more data,
which flows back in this loop, and in this happening across
the entirety of the business. This sort of frame, this was our thesis
when we set out to build everything
that we’ve built. If you lay this down on its side… Let me just put the pieces back together
now what Satya did this morning. First thing we did was lay down
a planet scale footprint of data center capacity, that today we enjoy
more regions in AWS and GCP Google combined. Massive scale. Exabyte class data storage to capture all of this
information flowing from everything. On top of that, the Azure
intelligence platform and application platform allows you to take that data, turn it into insight,
predictions, and then to act. On top of that,
we built the Power Platform, which allows not just
professional developers, but citizen developers
or anyone who wants a low-code/no-code sort of high productivity platform
to build applications to do that. On top of that, Office 365
and Dynamics 365, which is the industry’s
only comprehensive sort of left to right business
application family, unstructured business
processes with office, communicate, collaborate, create with dynamics
across marketing, sales, service, operations,
finance, talent, and commerce. It’s holistic across the business
and integrated. These are sister services
that were built to work together. Then finally on top of that, AppSource,
which is our ISV marketplace, where we’ve got tens of thousands
of applications now from ISVs that have built
on top of this. Now, the important part here is that
this is a unified family of solutions. The Power Platform that we’re
going to spend time looking at today
is the extensibility and customization mechanism for Dynamics 365
and Office 365. When you open up
a modern SharePoint list, you can turn it
into a Power App. When you’re trying to enhance
the capability in Dynamics 365 for sales, you can use Power Automate. Likewise, it is integrated
downward into Azure, and we’ll show that today. When you’re creating a Power App,
you never run out of gas. If there’s something that you
can’t quite express In PowerApps you can always
drop into Azure Functions because PowerApps is natively integrated,
they’re aware of each other. And so, this stitching together
into a unified platform is what Sayed was talking
about this morning. It’s so important because I think
regardless of where you are in your journey and whatever problem
you’re trying to solve, we’ve got the ability
to solve at high, but you never run off a cliff
with the Microsoft Cloud. Okay, so let’s now focus
on the Power Platform. So, today we announced the name
change of Microsoft Flow to Power Automate and we also added into Power Automate
robotic process automation UI flows, the ability to not just with Flow, interact with APIs,
but also to interact with UIs and by the way,
in the same exact framework. And so you can pull these systems
together whether they’ve got programmatic interfaces or not and automate workflows. We also announced a new product that we demonstrated inside
this keynote, Power Virtual Agents. It allows you, without writing code,
to build incredibly impressive virtual agents and we’ll demo that and we’ll talk more
about the capabilities there. But we’re going to walk across
all of these capabilities today and we’re going to do it
in four sections. So for the next, call a hour or so,
we’re going to talk about how the Power Platform enables an organization to really go knock down
the app backlog to digitally transform by empowering those who are closest
to the business processes to go build applications,
automate processes. We’re going to talk about
governance and administration and how, unlike in the past
with a lot of these citizen platforms where you ended up with app
sprawl and shadow it, that you get the best of both worlds
with the Power Platform. We’ll talk about the role
of professional developers because you still need them. In fact, you need them
more than ever before, and we’ll talk about
the intersection between the Power Platform
and professional developers using Azure. And then, finally, we’ll have
Amir Netz show us some of the latest and greatest capabilities
with Power BI and talk about the data that’s being created
through these applications, how you can get
more value from them. So the first stanza,
or the first section, we’ll talk about unleashing
your organization. And I’m going to do this by chatting
with Ayman who’s the CIO of TruGreen. Ayman, why don’t you
come out and join me? Please welcome Ayman. -Thank you.
-Thank you very much. -Thank you for having me.
-Yep, thank you. So, I think we met each other
for the first time in April. -Yeah.
-It’s amazing how far you’ve come in just a few months. -Exactly.
-First tell me about TruGreen so people understand. -Sure, sure.
So TruGreen is the nation’s largest lawn care company. We operate also in parts of Canada.
We’re about $1. 4 billion. We provide services from lawn care
in the yard, trees and shrubs, and some pest control as well. -Fantastic.
And we show, I think, in Saeed’s keynote. -Yeah.
-So tell me a bit about what you’ve built. You’ve harnessed the Power Platform. I think you’re using it pretty holistically. What are you tackling? -Yeah, so it all started for us
early in this year when we started working with Felipe
from the Power Virtual Agent. And at the time we were set
to build a chatbot to help our customers check
on the account balance, check on when is the next service
reschedule, and so on and forth. And soon after, after that, we met in April
and at the time James introduced me to the catalyst team,
Jeff Holland, and the rest of his team. And really we started to take a different
look at the Power Platform. This was really our first introduction
truly into the Power Platform and it was very clear for us at the time
that the approach that Microsoft took, it wasn’t necessarily the hypotheses around
how you build an ecosystem. It really took the theories out of it, took it really into reality as we started to learn more
and more about the Power Platform and, at the time, I think we had just finalized setting
our digital transformation strategy. And looking into the Power Platform
and starting to work with the team, it did help enable
and unlock a lot of the capabilities that we envisioned to do
as part of that platform. -Got it. Now I think one of the things that has been
most impressive, watching you operate, is just how quickly you’ve been able
to go service the business. -Yeah.
-Talk about the velocity. -Yeah, so just to give you a few examples. And this is, like I said, we met back in April. Beyond the Power Virtual Agent we have a few examples
here to share with the team as well. As part of the Power Platform
we built an image recognition capability and it’s based on cognitive vision,
which is part of the Power Platform, and this capability is really a function
out of our consumer mobile app. You literally pointed at an object
in your yard that you don’t have any idea what could that be,
whether it’s weeds or insects, or what have you, and it pulls out the relevant information
through the AI algorithm in the background, and the database
on Azure Data Lake. And it would suggest for you
the treatment plans, it would suggest for you if there is anything dangerous
about touching that kind of object. We’re actually, right now, we are in the process
of looking at using the same capabilities plus cognitive deciphering, if you will, of information or of data
with the Bing Map API. So, as you probably know, if you would call us,
what we would do is we would draw a map
around the surface area based on our satellite image
that we pull from Bing API. So, what we’re trying to do is we’re trying
to take the guesswork, we’re trying to stay away from us
trying to measure it polygon on the screen, and we’re also trying to see
if we can take that image, satellite imagery, and be able to measure
the surface area on our own. -Fantastic.
-Yeah. There are a lot of examples here. We’re even taking phishing scams
on emails as well, we’re taking it a little bit further beyond a classic
ora traditional solution as well. So, we’re trying to scan our emails
with some of the cognitive services as well to detect patterns that are not part of
a wide massive industries, so really a lot of exciting things… -I love your passion on this stuff.
-Yeah, Really, really great stuff
and thank you for your support. -Yeah, yeah.
It’s like every time I talk to you I think I learn three new things
that you’re building. -Yeah, yeah absolutely.
-Great. So look, thank you so much for joining us. -Thank you.
-We’re going to have Charles Lamanna come out. He did some- -Thank you very much.
-Thank you very much Ayman. Thank you. So, Charles Lamanna did a demo of some
of the TruGreen capabilities in Saeed’s keynote. We’re going to do the unabridged version here and show some of the other things
that Ayman just talked about. So Charles, come on out and show us
TruGreen’s example. -Thanks James.
-Thank you. -I think the biggest addition this time,
as you’ll see in a second, I have a pretty
awesome stage prop. -Awesome.
-So you’ll see, we’re going to use that in a second. So, as we saw in the keynote,
there’s a lot of great capabilities in terms of how TruGreen
uses the Power Platform. On their website, they use
our virtual agents so you can go and actually report the issues
that you’re facing when you want to go make a reservation
with one of their field agents. So I can do things like I have
a lawn problem, identify the same types of issues that we saw
in the keynote are related to the different AI backend if it were the most likely reason
that you’d have a lawn problem. I can say, “Click through this,” and have a very natural
conversational experience, see more information
from their knowledge base about what can cause
those brown spots, and even actually click through
and ultimately schedule an appointment. And all of that able to happen inside of
that Power Virtual Agent experience. A simple low code no code offering
to be able to do everything from greetings and looking up information all the way
to scheduling that appointment. And the way that our virtual agent
does that is with a visual designer, just like you know from the rest
of the Power Platform. We have a great representation
on the right-hand side over here, so you can see all the different pieces
of information used inside of that chatbot. And if I go walk through the same
set of information that the customer would do on the homepage,
I can actually use tracing, which is… to go see… usual steps and how AI is deciding
which direction to go in the chatbot as well as look up the relevant information
to show inside of that experience. So if I click through that,
we can see those same type of options are being pulled all very naturally inside of
the Power Virtual Agent experience. And what’s great is that the Power
Platform works well together. So in the case of TruGreen,
they use Power Automate to go integrate
with their scheduling APIs. So I’m able to say,
“Look right here,” and that will automatically go out to use over 275 data
connectors available out of the box, so I can very easily schedule
an appointment for at the right time for the customer
in the right place. And that type of connectivity
is what makes it possible to have that very rapid transformation that we see
so often with the Power Platform and it’s a very short period of time
you can get a project off the ground. And the really special part
of this chatbot is what it does with Power Automate’s
brand new UI Flows capability. So, for this next piece where I want
to go record the fact that I booked an appointment, I can do that using UI automation
completely end-to-end. This new robotic process
automation capability inside of Power Automate makes it so that you can reach every
single application in your IT estate. You no longer are bound by the APIs
or data connectors that we ship. If there’s a UI for it,
which there probably is if a human uses it,
you can go use the recorder … quarter to actually kick through,
all those different details, and not have to worry about
how you integrate a legacy system. So if I say, “Come here and tick the email
input that was provided by the scheduling API,” all is available right inside
the TruGreen application. When I do a search, that will go and look up inside of the contact
management application, so that it’s the relevant
information, the activities, and actually paste in all the quotes
that are captured by that app. And it’s really important
because when an agent goes out to the home, they’re now able to see on tax
that the customer provided, inside that TruGreen website
and inside our power automate, this is the same visual designer that you already know
and love from Microsoft flow, except in addition
to the API calls or the AI integration
that we already have, you can now use
that UI flow capability to see all the different mouse clicks
and keyboard strokes that you’ve entered and actually trigger
those in the background, or in the foreground,
using these new UI flow capabilities. Just like that, once I confirm
that appointment, we fire off that
UI flow that I built. And what will is it’ll launch
the same win-32 application, host on the same server, it’ll go enter the email address
that the user provided. It’ll go do a search through
that text management application. I don’t do anything on the
computer the whole time… all the relevant notes. Thank you. This is only the first part. There’s actually three parts
to this story. Once you actually
make the appointment, TruGreen has to send
an agent out to your home. For that, they use power apps
on a mobile device. What this power app
allows their agents to do, is I can see all the different
appointments that I have coming up, and go drill into them to see
where does the customer live, what TruGreen plan
do they have? And one of the most compelling
parts of it is how you can use AI. Using object detection
and AI builder, TruGreen has made it
so that their field agents can use this power app to do things
like detect issues on lawns. If I go zoom into this photo, -take a picture of what those spots are. Don’t know what those are,
but hopefully AI will tell me. This will go up to the cloud
and it will go look across a bunch of images that they use, and it will tell me that there are
two necrotic ring spots on the lawn. All of this is done in an AI
and completely automated fashion making every single triggering agent
better at doing their job. That’s what technology
is all about, going and transforming each
different piece of your company. If I click through here, I can also see
there’s additional information about the necrotic ring spot. If I want to read up about
what exactly it means, it seems it’s extremely serious,
so I should be concerned. I can then go click into a live report, which will show me all the other instances
of this particular issue in the area. I can be informed about what’s impacted
all customers I may serve as a TruGreen agent. What’s great is I can go report,
record additional notes, and I can say something like, “this grass has
two necrotic spots”. This is where something’s
really interesting. When I click this button
over on the computer, we’re going to be able
to see a UI flow trigger. If we can proactively switch
over to the computer. Great. Once I submit this note,
it actually is going to use that same UI hole, that I just built earlier,
because not only can you integrate your chat bots with…
capabilities. You can also wire up power apps
in any other system even running in the background
to do the same type of thing. What it does is it just goes
and triggers that same UI flow that I was showing earlier
to run on this machine. I think it’s actually on the other
computer, but that’s fine. But so if we look here,
we can see that the next step of it is how that power app was built. That was built inside
a power app studio. For those that aren’t familiar,
a power app studio, is just a mix between two note
from PowerPoint, along with Excel. A very easy way to go create a web
or mobile application. Though if I were to click
through these items, it would work just as it did
on the iPhone device, and just as a would
in the Android device, right here in my browser,
without having to rebuild or recompile. We can do things like that. Identify concerns, control. This is that AI builder… detection control. We can see it right up here, and I can go select
whatever AI model I want. In this case I trained it,
to detect common lawn problems, but you can go… it for a whole bunch
of other different topics, and all kinds of other field
service and user scenarios. This intelligence embedded
in the application, is what’s so important about the way
we look at the power platform. A modern app, a modern dashboard,
a modern automation or a modern chat bot, all needs to be infused
with intelligence. Otherwise, it really isn’t
a modern experience. That’s a big focus for us
across the board. If I go down to this live report screen,
we can see that on this submit button, this is actually calling
that Microsoft flow to go run that UI flow
that was configured earlier on. All of this is wired up in a
very consistent, cohesive experience. And that integration across
all the different products we have in the power platform is what makes it really
differentiated from a low code environment. It’s not just about apps
or anything else. You can bring it all together
in a seamless experience. To show that AI builder model, this is just all you have to do
to go upload those images and train it, so you’ll be able to go configure AI
and across all of your applications. That was the second part
of the three part story. Once you actually have users
who go and say get your TruGreen services. The next step is you want to go record
how they pay you through invoices. And the way that is done is probably
the most common form of business to business communication,
which is an email with an attachment. In this case this is
a background power-automate step, which has a trigger for
whenever an email arrives. This trigger is using the office
365 Outlook APIs integrate. This runs entirely in the cloud. Based on the attachment,
we’re going to run a prediction on it and the prediction is going to use something called
“forms processing” from AI builder. And this is going to allow us
to take an opaque JPEG, PNG, or PDF file and actually extract all the relevant
structured information from that form, and show it inside
of the design experience. To show how that’s done, I’ve already
uploaded five example forms, because in all the cases of AI builder, you don’t have to go get Python or be a data scientist,
or understand machine learning. You just need to go bring examples
to us, highlight the parts that you care about,
and you can go actually leverage AI. With those five forms that I’ve uploaded, each of which looks approximately
something like this… can just click next, click analyze here, and we can see, I can highlight
right on this screen in the browser, all the relevant pieces
I care about. Maybe I want to capture
the invoice number, I want to capture this table,
which has the lawn items. I want the balance due,
and I want the due date, and I want the total. And that is all it takes. You upload those files, and you can just
go extract all the intelligence from it. Now that I’ve created that model, that’s what we’ll actually run inside
of this power automate example. After I’ve extracted
all that information, I can do things like post it to teams,
to get an approval process, so humans can know and look at it
before it moves forward, then I can go send an email. I Can even also add things
like going to UI flow, or add your services all from
the same powerful visual design surface, which we see consistently
across the power platform. That’s again, just the biggest
takeaway I would say, from this demo, which is the ability to weave
together all these different capabilities. Power virtual agents, power apps,
AI builder, power automate, and the new UI flows together,
make it possible in a very rapid fashion to transform
how your business operates. And after that I’ll hand it back
over to James, but thank you
for the time today. -I think it’s just a great example
of how all this stuff fits together. Super synergistic
across all of these services. They were made to work together, and when we introduced the virtual
agent capability, it was stunning. In fact, what customers
started to do with it, this ability to tap into power automator,
flow, as it was known, from the virtual agent. You’re not just talking with a customer,
you’re talking and acting on behalf of. It’s just a super powerful, one of the things
that’s probably non-obvious, just to go back to the synergy thing,
is when you build a power of virtual agent, when you are providing keywords
that you’re going to respond to what the customer will utter
when they’re interacting with the bot, you only have to give a few examples. I would’ve thought, “Hey, it’s rule-based
and here are all the things”, but no, you provide a few examples,
and because we have Bing, we understand
how people express things. Billions of searches
where we understand synonyms, we understand
how humans ask questions. We’ve taken that and we’ve infused
that into power automate this hidden capability
based on these. They’re just very powerful,
and I think really shows how all of this comes together to create something that is far, far greater
than the sum of the parts. Now, trying to be funny here,
we talked about unleashing the organization. If you look across organizations
that have adopted power platform at scale, hundreds and thousands of applications
and dashboards and reports and flows are being built,
it’s just massive, adoption and scale that happens pretty quickly
when it lands in an organization, and that could be
terrifying, frankly. I’m an old guy, and I started back
when dBase II was the tool of the day to go do low-code/no-code,
but we’ve seen it over time, FoxPro and Access and all of these platforms
that allowed non-developers to go, in very sort of high productivity terms
to go build applications. The challenge was, in the past,
they were often running on boxes sitting under someone’s desk. Someone would kick the power cord out. That person had left the company. You never understand what’s happening,
and it really creates a nightmare from a governance
in a management perspective. With the Power Platform,
you sort of get the best of both worlds. You’ve got the ability to,
like we were saying, long-leash your organization. You’re giving the organization
plenty of room to go innovate, but you’re still governing. SNCF is one of the customers,
so SNCF is the French National Rail System, and they’re global
transportation company, one of the early sort of big adopters
of the Power Platform. Early on in their journey,
and we’ve seen this happen now company after company after company
where the Power Platform gets in, and it starts to spread, and then the IT organization
comes over the top and says, “Great, let’s get
our arms around this.” They do two things, typically,
“Let’s get our arms around it and show that
we’re governing correctly, and let’s go empower
the organization. Let’s create a center of excellence. Let’s go train people, and let’s
really get the most out of these assets.” To show how they’re thinking about it, and frankly, how they’re thinking
about the role of it, let’s roll the video from SNCF. -[Speaking French] -[Speaking French]
– [Speaking French] -All right, so you may have seen
one of the applications pop up there. It’s called Digi Bogies. It’s my best French impersonation, and a bogie in French
is sort of a chassis, a train chassis. It’s the axles and the wheels and the… you made it with a tank or a boxcar
or something like that. They’ve got a depot where they’re constantly
sort of creating these cars, putting them together. There’s no app for bogie manufacturing
that you can find on the app store. The person who’s in the depot, the person who actually does the work built
this application that allows them to go from… and these numbers are incredible,
but they say they’re true. It’s hard to believe. They went from a 90% failure rate,
so 90% of the time, they had to try multiple times
to get the right match between a chassis and the car, to a 20% failure rate by using AI Builder
and using a PowerApp to do the measurement. It’s stunning. I think if you think about the applications that are potentially most valuable
to an organization, it really is the apps that you can’t buy,
you can’t find in an app store, and the Power Platform
allows you to go build those, and that’s why
all of these opportunities that are presenting themselves are being knocked down
by our customers using the Power Platform. To show a bit about how to manage
the applications, how to govern
the Power Platform environment, I’m going to have Ryan Cunningham
come out and demo those capabilities. -Thank you, James.
Hey, everybody. You might be surprised to find
if you’ve been following us for a while that I am not going to show you
how to mate a chassis to a train car today or build an app to do that
because at the scale that SNCF and many organizations
like them are operating, it is no longer about one app. It’s not about building an app
or a business process a little bit faster. It’s about hundreds of applications
for thousands of users. At some organization, an increasing number, it’s about thousands of applications
and tens of thousands of users. Now, sitting in a room like this, that type of scale is going to
give you one of two reactions. Either you’re going to be thrilled
that it is finally, finally pragmatic to start
achieving digital transformation at scale in organizations, or to James’ point earlier,
you’re going to be terrified. Who’s terrified? Be honest.
Yeah, I know. I get it. But before you go running
for an off switch, hear me out. I want to demo to you the most important
feature of governance and administration
in the Power Platform. I’m going to do it
in a single click. You ready? Bam. Now we chuckle, but here’s what sets
Power Platform apart from every other
low-code option of the past, every other shadow IT
or point solution SaaS provider or other thing that’s happening
inside of your organization already. Everything I do on the
Power Platform as a maker, as a user, all governed and authenticated
through Azure Active Directory, planned at scale, enterprise-grade,
identity management system, and that means that
this is not in a shadow, it’s not on a box under a desk, it’s not sitting
in a file on a local drive. Everything I do is inherently
governable by default. Automatable, auditable,
manageable, and all of that starts ground
zero for governance of the Power Platform starts at the Power Platform
Admin center where we expose a lot of that visibility just right out of the box. What resources am I utilizing
across organizations? What are people doing with PowerApps? What exactly are they making,
and who are using those apps? All of that data exposed here,
right in a few clicks. I can go all the way drilling down into what apps are throwing errors
and how to debug them. I can go even further than that. I can look at utilization inside
of the common data service down to the API call volume,
all of that exposed right here so I can take action on it
and respond when I need to. I’m always acting as myself. I can never do anything as a user
through this platform that I couldn’t otherwise do
at the front door of a service, and all of that is governable
here in the Admin center. But you can go further
even than that. You can start to set data
loss prevention policies across the hundreds of connectors
that the platform exposes, and there’s
a lot of services here, but I can come in
and create environment-specific or tenant-specific data groups. I can add SQL Server or Salesforce
and say, “Look, this is for business data only. Never let data from
these particular sources go to other sources
in the platform.” I can get even more granular than that. The manageability is just built
with tooling here. But that’s not all. In fact, that’s just
the beginning. See, managing and governing
the power platform at the National Railway in France is different
than what it means to manage and govern the Power Platform
at TruGreen lawn care or at the Tacoma Public
School system or at any of the thousands
of other organizations that are running
PowerApps today, and so governance
is inherently extensible as well. We know extensibility here. We know how to do low-code extensibility, and in fact,
we’ve connected the entire platform to the platform itself,
which is a bit meta, but it means I can use PowerApps, Power BI,
and Power Automate to manage PowerApps, Power BI,
and Power Automate. I want to show you some examples
of how SNCF are doing exactly that inside
of their organization today. You saw a hint of this
in the video, but let’s watch. It all started with one app,
one maker, one user. You’re watching 18 months of data
of the spread of PowerApps and Power Platform inside of this organization. In a custom BI dashboard that SNCF
are able to build and use over For our admin
and governance APIs to look at how adoption is progressing
inside of their organization. Now, this scale is staggering. Look at it spread across locations,
across individual makers, across applications. But again, this is not
the spread of shadow IT. This is a success story because IT is essentially
not only governing this, they’re encouraging it. They’re serving the needs of users
and the needs of their communities and upskilling
their workers as they go. They’re not just watching it
passively in BI, they’re creating custom apps
that go further than our out of the box experiences and show end users
exactly what apps they need to find. We talked about bogies earlier. Let’s go see if we have an app for that. Sure enough, we do. Not only are they tailoring
experiences to end users, they’re also automating
the governance process for makers as an individual app developer,
a citizen developer, and a business unit. I can go see exactly what I’ve made
and what type of impact those apps are having and I can interact with the IT
in an automated fashion. I can go request an environment for a particular app
and fill out why I need it, maybe fill out
the location of my database and if I need CDS,
I can say if I need advanced capabilities and go off and submit
this to central IT for approval. All of that automated in the back end
using power automate. I can set up custom governance processes
that fit the needs of my organization. I can say, “Hey, whenever one of those requests
comes through, if and only if they need
the advanced support capabilities, let’s kick off an approval process. Pass the details of who’s
requesting it and what they want and why. All the way downstream.” In fact, I can even embed
that approval process right into where
my governance team is working. Let’s say right here inside of Microsoft teams,
there, it just came through, this approval request that I made and I can approve
it right in line and move on. All of that folded into the fabric
of how we’re already working, making it possible to manage hundreds of apps,
thousands of users, hundreds of makers at scale. Now, that’s great for the long tail, but what about when I want to run really mission
critical workloads on Power Platform, what about when I want to roll out
a single app to thousands of users and manage it in a professional DevOps process that I’m used
to inside of my organization? Guess what?
I can do that too. Azure DevOps now supports build tasks
and automateable tasks for a whole range of actions
across Power Platform. I can come in here and not only learn how to do
DevOps at scale with Power Platform, I can use source control
to actually manage my code, I can dive in here and look at the output
of individual customizations and solutions, I can even diff that in source control
against previous commits, I can automate pipelines
across how to build either development or exports from my development environment
or builds up into my production environment and I can start to automate a whole range
tasks in DevOps across that spectrum. Let’s just search for a few of the Power
Apps tasks that are available here in DevOps. You can see it’s a host of things
from creating environments, packaging solutions, even running deep automated quality checks
across that whole spectrum of capabilities. All of that from citizen development,
all the way up to professional development, running more efficiently in low code. That’s how SNCF and again many organizations like them are able to scale this out,
a whole spectrum of applications. Many of which are used in individual departments
by citizens with a small numbers of users and are passively governed through data
loss prevention and through monitoring and some of which who are centrally
controlled and rolled out at scale to large volumes of users
by professional teams. It’s the mix of those two things together,
professionals and citizens working in concert that’s really
driving this transformation. The IT team on Rippedoe, the CIO,
they’re not catching up to this, they’re leading it. They’re actually creating
training programs inside of the organization to actively train
and encourage people to build these skills, to bring forward use cases that never before would have
even come to the desk of an IT person as a request, let alone get funded to be built
in traditional app development. That’s the important part of this. The scale is important and you
don’t get to transform thousands of business processes a year
with business as usual. It doesn’t happen with a few
more software engineers. It doesn’t happen with one
more integrator contract. It only happens when you start
to think proactively about arming citizen development
and Power Platform in your organization. It’s working for SNCF, it’s working
for hundreds of other customers and we’re thrilled to see
what you can do with it as well. Thank you very much.
Back to you, James. -Thank you, Ryan. You know, Saad, this morning in his keynote,
talked about half a billion applications that will be built over
the coming three years, more than all the applications
that had been built over the last 40. This is, in part, how it’s going to happen. You’ve got the ability here,
whether you’re a professional developer or not, or a citizen developer to
with very high productivity, go build these applications in a way
that was previously not possible and certainly not governable. We think it’s a very powerful platform. SNCF is a perfect example
of how to do it really at scale. We’re going to shift now
to the third topic. Sort of the role
of professional developers. I’m going to, in a second,
bring out Ladd. But Ladd, who is a professional
developer at Schlumberger, I met not too too long ago,
about a year ago. He is our biggest Power Platform
advocate at Schlumberger, professional developer, was very much not impressed
with the Power Platform or Power Apps sort of
before he took a look at it. To talk a little bit about the role
of Pro Dev in the Power Platform context, I’m going to have Ladd come out and join me
and have a conversation. Please welcome Ladd. -Thanks for having me.
-Thank you, Sir. First, talk about Schlumberger. -Schlumberger, we are the world’s largest
oilfield services company. We have 100,000 employees in 86 countries and we use a combination of technologies
like engineering, manufacturing, and a lot of computer science that other companies
like Exxon and Shell generally produce oil. -Got it. Before we get to the Pro Dev thing, so we just came off
of the governance side. I know that Schlumberger has
thousands of Power Apps. How are you dealing with
administration and governance? -Sure. Our Power Apps journey
started early last year and really, it was a way for us to see if we could get our employees excited
about building their own stuff. We are a company of engineers
and scientists with a great hunger for technology
and a lot of shadow IT, right? We really wanted to find a way
to make them digital. Flash forward,
over the past 18 months, we have over 9,000 Power Apps, 11,000 Flows and countless BI
reports created- -That’s amazing.
-… inside of what we call
our innovator environment. What is our innovative environment? This is a place where we can finally
let people, we used to call it shadow IT, give them a place to play
to really see if their ideas have value, but also has some guard rails in place, so we’re not
tweeting out intellectual property. -Got it.
-This is really important because we have to be able to provide
some IT governance if we’re going to have applications
mature at the same level, but provide some freedom to let these,
our citizen developers try and build. -Fantastic.
Let’s get to the other topics. Do we even need Professional Developers
in this world that we’re sort of entering? -It’s a great question and in fact,
there’s a lot of fear, I think, on low-code
with Pro Dev for this reason. Their answer is we needed even more
if we’re going to do what Saad, he talked about being a tech intensive company,
we have to help our employees navigate. Let’s be honest. It’s hard for IT to navigate IT.
-Yeah. -We have to provide, I think, first of all,
a nurturing environment to let our citizen developers connect, be together, get advanced,
get advanced training, share best practices. But what we can also do
is one of the biggest questions we get in our COE is provide access to data. -Yeah.
-Data that seems to be locked away in our legacy applications in chains sometimes,
but also give access to advanced, the development functionality like Azure Function,
which you’ll hear by here in a minute. -Yup.
-But finally, we should do what we do best. We should deploy applications
at scale. This really is powerful
with the Power Platform because you can take these small seeds
of innovation that happen, let’s say, during your employee’s
daily life or through a hackathon, which we’ll hear by here in a minute,
and scale them globally, really 10X their value. This is really what Pro Dev teams do well. For us, you’ll hear about an application
called Get Paid. That’s focused
on accounts receivable. Right?
A problem we all have. The app was built over 2 days
in a hackathon with 1,000 people participating and it was
the number 2 winner, actually. Now, we’ve taken it centrally,
we’ve provided access to data through data lake and through a couple of
Azure functions, as well as some APIs. Now, we’re going to scale it globally
and we’ve actually been asked to accelerate it in the next two weeks
to hit the books now. -It’s great story of taking
something small, a small seed of innovation and delivering it at scale
through professional development. -Fantastic.
Ladd, thank you so much. -Thank you for your anytime.
-Thank you. Thank you. To have a look at the application
that Ladd was talking about, I’m going to invite
Julie Strauss out to do a demo. Welcome, Julie. -Thank you.
Hi. -So enabling people to innovate
is the key tenant with Schlumberger. As you heard Alad, just say the ability
for them to enable all of their developers
to amplify their skills, come together with the business
and empower their makers to accelerate their application process,
is super critical for their success. So I’m going to show you
the Get Paid app that Alad talked about and this is really
a fantastic example of how the pro developer community comes together
with the maker community at Schlumberger, and they are really transforming
how they do business. They are generating
a multimillion dollar increase in the company’s
working capital by working together. And how did they do it? They modernized the business process of transforming disputed
invoices into paid invoices. They allowed the professional
developer to create APIs in the tool of their choice, expose it through
asher functions, into PowerApps where the makers can now
start to digitalize that process. So why this collaboration? You’ve seen how powerful
that all the makers can create apps in a matter of hours
using PowerApps. And honestly we are super proud
of what we can enable users to do without writing any code. But the reality is you cannot
do it all without writing code. You will hit a wall somewhere. And what we are doing, we are embracing the skillset
of the professional developer investing in tooling that will allow the developers
at Schlumberger as well as developers
in honor enterprises to create Lego blocks that they can now share
with the makers, with the citizen developers, allowing them to go beyond what we can do
out of the box with PowerApps. So I’m going to show you two pieces
of our extensibility story for the pro developer. First I’m going to show you
two labs point how we can take, allow the professional developers
to create APIs over legacy systems, transform that using asher functions
and expose that inside PowerApps for the users to get access to that data
that is really locked away. Then second we are going to say
as the maker create these applications, they might run out of steam
again in the UI layer if they want to pull in component
that they can’t really customize, that is not available
out of the box. So we will pull in again
upon the developer’s skill set to create custom controls. So let’s start here in visual studio
where you can see I have built my API that is connecting to that
legacy account receivable system. Now the reason I had to build an API
is there were no APIs is available
out of the box. There are no connectors, meaning the makers can never get to that data,
but you don’t want to move it. You want to keep it where it is. So by developing this API,
I really have the need of my pro dev skills and because of the deep integration
with asher into vs code, I can actually go ahead and deploy
this function directly from BS code. Now before we go there though, I want
to take a little look at the performance monitoring I’ve been doing with my API
inside the asher portal. And you can see here I might actually
run into some performance challenges for each of the get invoice API calls
is actually taking more than a second. Now that can be a challenge
to Alad’s point. This application will go out
to thousands of users, but it’s a bottleneck that can be really hard
for me to solve in the API layer. So that’s actually where the power
of the API management come into, into the picture. So I’ve gone ahead and I have onboarded
this API into asher API management
and I prepared it in here. So it can see here
I have my invoices API. And what I can do with API management,
I have a central location to govern and manage these APIs built
by my pro developers. And it’s not just this API
that I built for this specific task. It’s all of the API APIs that my pro
developers are generating for my company. So I have a central place to govern. What is particularly interesting in this case is that I would actually go ahead
and I will create a policy. I want to create a caching policy that allows me
to cast the response from the API to accommodate
for that deficiency. Then the performance. So I’ll go ahead and I’ll say I will cash
this call every 60 seconds. So the impact obviously as this API
is exposed through PowerApps, each of my users hitting it
will get an instance, reply back, response back. So I can go in here.
Now. Next thing I want to do, I want to export
this API or the function into PowerApps. And I can use the out of the box integration
we have enabled where I can export this Asher function into a PowerApps API. In fact, it will show inside
PowerApps as a custom connector ready to use. Now, before I go there, remember
Alad set again, key is amplification. So another part of API management
is allowing the developers to amplify their work
through the developer portal. So Asher API management has a developer portal
where all my work can be published. And as a developer I can make my work
discoverable for all my end users. So if I’m in PowerApps and I cannot
find connectivity to the data source I need, I can go out here
to the portal and browse and see if somebody’s already created
an API for me to use. Now if I was a really nice developer
and I believe I am at least today, I actually went ahead and published
that connector into power app so it’s already available for me
as a maker in my environment. I don’t have to go search for it. So it’s listed in here. So now if I change my hat and I am a maker,
I will go ahead and start creating the application that will allow all of the users
that has access to that application to easily see which invoices are due. So I’ll go in here and insert a data control here
so I can connect to my data source. I will just change the look and feel a little bit
to make it more compact and I’ll go ahead and wallah. This is where the custom
API done by my developer meets the maker. In this case. They come together right here in the UI and the user,
maker can start using and connecting to this
API to the data source directly. So I’ll go up here and I’ll connect
to the get invoices table, and you’ll see in a second
that I can actually go and I can get visibility
into the schema of this legacy data source. And I, which would make it easier for me
to decide which fields I want to dispose
and disclose in my application. So next one I want to go do. I want to go and connect up each of the fields
that I want to be working with. So I want to see the number of days
that each invoice is outstanding, and I want to see the customers. But what I also want to do, it might actually be useful if I can also see the amount that is due,
because that will help me prioritize. And all I have to do is copy a field and tie
that into another data field from the backend system. That was pretty easy, right? So now Schlumberger move from 80 days of email and Excel to figuring out which invoices overdue
to instant access to what invoices are overdue. So they moved from 80 days of data
collection to five days of dollar collection. So that’s a pretty good investment right there. Now you don’t have to stop here
because they have a custom connector. I can also employ the power of power automate. I can start be proactive. I can create flows that will call me and notify,
for example, the account team and say,
look, you have invoices overdue. Finance can react. Instead of going to the app,
they can receive notifications when invoices are overdue. So that’s powerful. So now you’ve seen how I could
connect to data from my APIs. Next stance is looking
at how can I extend my UI? Imagine in this example, it would be
very compelling to associate a peep, I person with each
of these line items. So what I really want
is a people picker. PowerApps doesn’t have
a people picker. But here again I can call upon,
this goes to my pro developer and they can actually create
a custom control. But what is even better,
we have enabled you to take custom controlled already built
and use them as part of PowerApps. So I actually have a office
fabric react control that is that people picker
that was already built. That I can now reuse inside
PowerApps. Actually let me
just cut this out, and go because I want to associate that
with my invoice creation here. So we’ll add my people, picker field. So this is the react control already
built for another application that I can simply reuse. And I can play the app
and I can add people here. Obviously I’m using sample data as pod. You want to hook it up to the graph so you can pull directly
in associate employees. You want to take the output
from this control and embed it into the form. Super powerful. So we already have a very vibrant community building controls that are out
there publishing their work. So talking about amplifying. So if you are a Pro Developer,
you are thinking about working with PowerApps, it’s time to stop
thinking and start doing. With that, I will hand it back to James. Thank you.
-Thank you Julie. So I think it’s fascinating too. We’ve been talking about service
oriented architecture for at least 20 years
at this point. And what Julie showed really was SOA. I mean, being able to expose
building blocks, managing those through API
management and creating a… She called it Lego blocks for
the citizen developers to embrace and begin using. Again, super sort of,
it illustrates the ability to take all of the capabilities
from Azure, from the Power Platform and to really pull them
together into a unified whole. So a lot of the PowerApps
that are being built are being built to replace paper
based processes. If you look at… So I was at Heathrow airport
just maybe a month ago and I was visiting with someone
who works in the security area of the airport where they bring vehicles
off the street. They bring them into
a secure environment, they inspect them and then
they let them into the secured area. That person, the security guard
built a PowerApp to automate. They used AI builder
to read the license plate. They automatically generate
the search areas. They have to randomly generate
three areas. And so he built this PowerApp
that took what was a paper based process that was rife with problems
from a transcription perspective. They had a lot of issues
about data quality. And because that PowerApp
is now capturing the data and storing it in the common data service,
it’s available for further analysis. And so Power BI really does
sort of complete the loop here. We talk about automate or analyze,
act and automate, Power BI, PowerApps, Power Automate. This analyzed part is critical,
but it creates this virtuous cycle. So you’ve got these applications
creating data, you learn something, you create
more applications and automation, it just really spins up. So to talk about Power BI and what one of our customers
is doing with the technology, I’m going to have Amir Netz come out
and do one of his fun demos. Going out. -Thank you James.
-Thanks. -Veolia is one of the customers
that we are working a lot with. If you don’t know Veolia from UK,
they are one of the largest waste water and energy management companies in the world
with over 150,000 employees serving hundreds of thousands of businesses
and millions of households every day. They are going to show
their own transformation and for the data
from the application, and business processes
are being streamed to CDS into Azure. And they’re using Power BI
in order to analyze the data, gain insight and optimize operations. What you see on the screen here
is an application that is very much like the one
that truly use internally. And this beautiful application
is branded with the Veolia colors and logos. It has this beautiful dashboard inside it,
but it’s not just dashboard. Veolia is taking advantage
of the fact that you can embed into Power BI applications,
other kinds of artifacts. For example, for new employees
or just being onboarded to application, they have this video that tells them
how to use Power BI. And the documentation of the
applications also available in here as a word document
just like that without any coding. And if the employee missed
the executive review, that PowerPoint presentation is
available directly within Power BI. But let’s go back into the report here. And you can see a typical report
we are tracking here the main KPI Veolia is concerned
about like diversion percentage, recycling rate and total weight. And you can see that we have
here an interesting trend. We have a decline in the total
weight manage by Veolia by month. There’s a decline here
from August to September. And as an employee,
I might be wondering why? And this is where the power of AI
and Power BI comes into play. You can just right click and say, “I’d like Power BI to explain
the decrease from August to September.” And just like that, Power BI
analyzes dozens of possible combinations and tell you what is
the most likely explanation. For example, in this case we can see
that most of the decrease is from the non-hazardous
industrial waste, but we can have other
explanation available as well. So that’s great if you have the data
you’re looking for on the report itself. But what if you have a question
that the report creator did not think about? This is where Q and A come into play. And Q and A, we just released
a new visual for Q and A, this natural language processing capability
just in the last few weeks. It allows users to ask in simple
English questions about the data. They get here a set of suggestions
auto-generated by Power BI so they can onboard easily or they
can start typing their own questions. For example, I’m going to ask
to show me total spend. And as I do that I immediately
get the answer. I can continue asking. I can say, “Okay, let’s break it down by place.” And notice two things. First of all, I’ve got the answer
but also you can see that we have place marked
with two red lines. That means the Power BI doesn’t
really know what place is. Nowhere in the Veolia data model
the workplace is being mentioned, but because Power BI integrates
with the office… we know that place is a synonym for site
and since site is part of the Veolia data model, we can solve this query. So we can go and select site
and the question is the same one. The answer is the same one. Now this is a Power BI visual this
Q and A, with this AI natural language. So it means that you can format it,
you can theme it and it plays with a cross
highlighting game of Power BI. So I want to see, which sites
are sending their weight to the NHI final delivery side. And I just cross highlight
and I can see that very easily. Let me ask a different question here. I want to ask, I’m going to ask,
sites by the lowest diversion rate. And as I do that, I can see that Westminster is the site that by far
has the lowest diversion rate and I might be interested
to understand why this is happening. So for this, we are really excited
to announce that next week we want to release
a new AI visual that is really specializing
in root cause analysis. And this is the Decomposition Tree. So the Decomposition Tree
allows you take a KPI and start breaking down
by its component. And then what we look here,
the total span, I’m going to first break it by site. Remember that we had this
Westminster site, this is the one. And now I can continue to break it
by all sorts of other dimensions and I can do it manually
or I can ask the AI capability of the visual to show me what is the next best breakdown
to explain where are we spending the money. So I’m going to go and ask for the highest value, and it tells me if you’re looking
for the highest value, the best way is to look at actual name. And that you can see that manual sorting
by far takes most of the spend and I can go ask
for another highest value. And again, Power BI analyzes and tell me,
look at the waste. And we can see that most of the spend in
Westminster goes from manual sorting of food waste. That’s kind of interesting and maybe Westminster need some help from us and this
is where we combine action with the insight. And I can go and press the action button. What you see here on the screen
is that we have embedded a full PowerApp application into the Power BI dashboard. Which means that now I can really drive
from the report action. For example, I want to click here on the food waste
and I want to fill in the site name. Going to be Westminster and a waste type. And I would want to start the process
reinforcing the work in Westminster. So I’m going to go in, type in my name and I’m going to notify the site manager
that, get ready we are coming. And we just started a business
process of reinforcing. Now what does it mean? Okay, so we’re going to do play
a little game here. I want you all to take your phones
and we are going to go and become
for a few minutes mock employees of Veolia. So take your phones and go to this website
and sign up to become Veolia employees. So this is demoignite. com.
Pull out your phones. I know you had them on your hands all the time,
people tweeting. And take your selfie, put your first name,
where you’re from and sign up. You will be assigned to one of four teams
and you’re going to be swiping and sorting the waste, the food waste
with the arrows on the screen. We’ll do it in a moment. As you do that, as you guys
are signing up and it’s easier. As you guys are signing up,
you can see that in real time, we’ll see the signups happening on the screen,
because it’s really important. As we are combining insight and action,
we have to make sure that the insights
are really showing it real time, what’s going on in the applications,
in the processes that you are having. So we’re investing a lot in insight and we can see that Chris is the most
common name for you guys so far. Dan is just as common.
Jason’s just with there. Most of you guys are coming from the North America,
some of you are coming from Europe. Can see the photos of you guys,
Chris and Diana, now three and three
and we can actually… If you guys want to go to the next screen,
we can actually see who’s the game here. As you guys start swiping
and we can see here that Miguel is… Oh wow, the backstage team is 46 wives. You will get exactly 60 seconds
to do the most wives. The blue team is leading Mitt Michael here is… Sarah is actually taking the lead here. Very good.
So we can see it real time. We can see who is winning here. The production of the teams
and what you really saw together here is a beautiful example
of the combination of AI and real time bringing through
the data platform and through the PowerApps,
Power Platform, insight and action together. Thank you very much. -It’s always fun to watch
Amir do these demos and so this sort of takes us
through the session. So we started with sort of unleashing
your organization, talked about governance, talked about the role
of professional developers and closed with data. Good solid run through left
or right, the Power Platform. If you’ve not leveraged the Power Platform,
if you’re not building applications using Power BI, using Power Automate,
I’d encourage you to give it a try. Incredibly powerful platform
and two advertisements. Number one, if you haven’t seen
the 2019 release plan wave to plan for the Power Platform,
it’s available at this URL and there is a free
certification exam offer here that the learn team wanted me
to put up in front of you. So take a snap of the QR code. And with that, thank you
so much for joining us. Enjoy the rest of your day. Thank you.

You May Also Like

About the Author: Oren Garnes

Leave a Reply

Your email address will not be published. Required fields are marked *