A Power Apps Portal for 1 Million Users in 100 hours with Mohammed Mohsin Khalid, Imperium Dynamics

Neil Benson: [00:00:00] Welcome to Amazing Applications. The podcast for Microsoft business apps makers who are building amazing applications that everyone will love. 

Welcome to the show. Hi, I'm your host Neil Benson. Our goal on this show is to help you slash your project budgets, reduce your delivery timelines, mitigate technical risks, and create amazing agile Microsoft Dynamics 365 and Power Platform applications.

This week, we hosted our first virtual events for the new Agile Applications user group. Thanks to everyone who joined us for the first two sessions. One was held early in the morning here in Australia for our members in the Americas who could join us in their afternoon. And the second was held in the evening, here in Australia at least, for our members in EMEA and Asia Pacific, who could join us in their mornings or afternoons. The registration process was pretty easy and my experiments with the tech using the Microsoft Community Teams tenant seemed to go well. 

But I did learn not to use animations in my slide deck when trying to use PowerPoint Live in Teams. Those animations don't seem to be working yet. And the slides just kept repeating over and over. So word of warning. 

Speaking of which Teams Webinars meetings have just landed in this tenant, hopefully in my own tenant as well. So perhaps we'll try that next time to see if it makes it even easier to register and send a follow-up to you after you join a session.

You might already know that Microsoft has launched their public preview of their User Group functionality for the Power Platform, Power BI and Dynamics 365 communities. Microsoft got some work to do to combine all those different product communities together into one single platform. But in the meantime, you can visit community.powerbi.com, or community.dynamics.com or powerusers.microsoft.com. And you can navigate from there to the user groups page, where you can explore all the user groups that are now available. And of course there'll be links to all of those in our show notes. This episode is going to be customery.com/036.

To get into the Agile Applications virtual user group, if you want to come and join us, I've got a shortcut. Visit customery.com/agileusergroup. That's the word customer with a Y on the end.com/agile user group. In our next live session, I've got an expert presenter on agile testing, so we can up our quality game when we're building business applications.

Okay. That's all the good user group stuff that's been happening recently. Let's get back to this episode. 

My guest today built a Power Portal to support a million constituents booking their COVID test appointments, and his team did it in less than 100 hours. So four days.

He's Mohammad Mohsin Khalid. And he's the founder, CEO and chief architect at Imperium Dynamics, a brand new Microsoft business apps partner that's already building an impressive track record in our ecosystem. Mohsin is the recipient of a FastTrack Recognized Solution Architect award, which is a recognition of the experience he's acquired from nearly 20 years building Dynamics 365 and Power Platform applications. 

His story of the COVID testing portal is a testament, really to the amazing things we can do with the right team, the right platform, and as Mohsin points out, a simple architecture. It's a great story. Let's meet Mohsin.  

Mohsin, welcome to the Amazing Applications show. I really appreciate you joining us. 

Mohammad Mohsin Khalid: [00:03:51] Thank you. And you, it's a pleasure to be here.

Neil Benson: [00:03:53] It's good to have you on the show. So always love talking to people about the amazing applications they built for Microsoft customers. And just really excited to hear about your story. You were telling me backstage about an application you built in five days. We'll get into that one as well, but I wonder if you could just take a moment to introduce yourself and then I've got a couple of introductory questions for you. 

Mohammad Mohsin Khalid: [00:04:13] All right, so hello everyone, my name is Mohammad Mohsin Khalid. I go by Mohsin or Mohammad, whichever one you prefer. I am currently the chief architect and founder of Imperium Dynamics. I lead the practice mainly focusing on Power Platform and Dynamics CRM applications.

Neil Benson: [00:04:35] Great stuff. So a couple of nice introductory questions for you starting with, what did you have for breakfast this morning? 

Mohammad Mohsin Khalid: [00:04:42] I had my favorite breakfast for every day, which is a Parata. Which is just like a flat bread cooked in oil and and a nice egg, and cup of chai. 

Neil Benson: [00:04:54] Wow. That sounds great. 

Mohammad Mohsin Khalid: [00:04:56] Yeah. 

Neil Benson: [00:04:57] Nice healthy breakfast. 

Mohammad Mohsin Khalid: [00:04:59] Not that healthy.

Neil Benson: [00:05:01] If it's fried in oil, maybe not. 

Mohammad Mohsin Khalid: [00:05:04] Yeah, I just put olive oil on to, to make it healthy.

Neil Benson: [00:05:09] Good for you. Tell us about like the first job that you had and how you got that job. 

Mohammad Mohsin Khalid: [00:05:13] I got my first job right before I had even left my college and it was with the second largest B2B portal in the world at that point. So we were competing with Alibaba and I built out a trade messaging platform for them.

Neil Benson: [00:05:31] A nice way to start your career working in a huge B2B portal. Good for you. 

Mohammad Mohsin Khalid: [00:05:36] Yeah. They're still using the code.

Neil Benson: [00:05:39] Is that a good thing? 

Mohammad Mohsin Khalid: [00:05:41] Yeah. I have thought of that.

Neil Benson: [00:05:44] I would hate to think if anybody who's relying on the code that wrote whenever I first graduated. So 

Mohammad Mohsin Khalid: [00:05:49] Well pieces of it. Sure. Of course.

Neil Benson: [00:05:53] And tell us about your current role. Imperium Dynamics, I believe, is a new startup for you. So congratulations on starting your own business. 

Mohammad Mohsin Khalid: [00:06:01] Thank you. Yeah. Yeah.  Imperium Dynamics not been around for a long, just started this year.  I think we have an amazing start, as I see it. We're focusing on mainly the Power Platform and Dynamics CRM applications. I founded it about two to three months ago and based on just the whole COVID situation that  ramped up digital transformation and had us doing it in days, like one of the projects we're going to discuss here.

That initiative, what we're trying to do is actually fill the gap of practices and very well-versed practices within the Power Platform and the services that Microsoft business applications has to offer.

Neil Benson: [00:06:43] That's great. I think it's always amazing to see that people who are brave enough to start their own thing, grow it, train some new people, bring some new talent. Into our ecosystem. And then, you know, some of them get acquired. There's always lots of acquisitions going on in our industry as well. So good luck with Imperium Dynamics. I hope it really takes off brings you a lot of success. 

Mohammad Mohsin Khalid: [00:07:02] Thank you.

Neil Benson: [00:07:04] We invited you on to talk about one of your recent projects for a Microsoft customer building an amazing application. I wonder if you could tell us a little bit about the customer, if you can. And some of the background into them. 

Mohammad Mohsin Khalid: [00:07:15] Yeah. The customer specifically, it was a COVID project , I want to put it out there. So not your normal implementation. So this might, the timelines or the situations that happened in this project might shock a lot of people, but yes, in COVID times everything was new. So the customer itself was a state which has millions of people that are going to use the application, the Power Platform itself, for scheduling tests as soon as the COVID hit. That's kind of the background I'll give you from that.

Neil Benson: [00:07:50] Okay. So you're trying to schedule all the COVID tests for people to be able to come in without waiting too long in line. They can come in and their scheduled time and take their tests. 

Mohammad Mohsin Khalid: [00:07:59] Absolutely. Absolutely. So the state tried to build an in-house thing in two days with custom .Net application. Realized that it's not going to work because a million or more people are going to try to get an appointment for a test. And this is April, very start when the lockdown wasn't even in place.

So very new. Everyone just going in panic mode on how they're going to support their citizens. So this state approached Microsoft wanting to see if Power Platform could help them out.

Neil Benson: [00:08:32] Okay.  Where they already a Power Platform or Dynamics 365 customer? Did they have a lot of experience with Microsoft business applications within the state's IT department?

Mohammad Mohsin Khalid: [00:08:41] No, they were not. But they were an Azure customer. 

Neil Benson: [00:08:45] Okay. 

Mohammad Mohsin Khalid: [00:08:46] Yep.

Neil Benson: [00:08:47] All right. At least they weren't a Salesforce customer. That's a good thing. 

So you got a call to say, come and build us a scheduling application to schedule appointments for thousands or hundreds of thousands of citizens. How long did you have to build the application? 

Mohammad Mohsin Khalid: [00:09:01] We got a call from Microsoft Thursday. And Friday, we had a meeting with the internal team on what we can propose on this sort of request. But the customer came in and wanted the solution Monday night. So we basically had, I don't even say days anymore. I count hours. We had like less than a hundred hours till Monday to get something, you know, that an absolute MVP that is scalable and that can support features as the pandemic progresses .

But key criteria. It has to be scalable, has to be there on Monday night for people to start scheduling things on Tuesday.

Neil Benson: [00:09:51] So I've had sales teams come along and say, Hey Neil, can you build a demonstration application for us? We've got a meeting on Wednesday, right? So you've got two days to scramble and build an application. But yours went into production. Is that right? On Monday after really a one weekend of development. 

Mohammad Mohsin Khalid: [00:10:11] Yes, absolutely. Because it was that situation and we realized that it's going to be above a million people that are actually going to be impacted on whether this works really well, or it works really bad and it's going to actually impact lives. So we put together a team of 30 people. And I led the architecture myself to make sure each line or each service that we are going for is absolutely needed and nothing is an overkill in the application.

I think what led it to a successful go live was the approach of making sure that we are keeping it bare bones and not including things that will either delay or make the process of adoption and go live more complex and will require more training. 

The first proposed  idea for resource scheduling, you know, the Dynamics Unified Resource Scheduling was down by myself. Nope. You know, like that's just a training nightmare.  

Just to give you some more background before I go into that.

What the state planned was that they were going to convert the DMV site. In the U.S., or probably in Australia as well, you have these tests emission sites for testing car emissions, right, vehicle emissions, and they have four or five lanes. Every state has a multiple of them. So they will go in and convert those vehicle emission testing sites into COVID testing sites. So that solution, which would have people supporting on ground using some sort of applications to validate when people come in and then we just open to the public for the scheduling piece of it, like a website or a portal. That was the kind of need. 

Neil Benson: [00:12:02] So instead of driving up and they stick something into your tailpipe, you drive up and they stick something into your nostril instead. And 

Mohammad Mohsin Khalid: [00:12:10] And you stay in your car, you don't get out. 

Neil Benson: [00:12:14] Right. They've got all the infrastructure, all the lanes are there. There's the stations at each on each lane. Yeah. Good idea. 

Mohammad Mohsin Khalid: [00:12:20] And I think that's probably the first time in my career, I worked for straight 36 hours as well. 

No sleep, but yeah, that was the need of the time. And Yeah. That's when I chose Power Platform, pure Power Platform services for it.  We were like, okay, Power Apps portal could do this thing.

And I don't know if you have interacted much with the Healthcare Accelerator that Microsoft had. It's deprecated now because of Microsoft Health Cloud, but last year the Dynamics Healthcare Accelerator was available on Power Platform. So you could install it, just bare bones with healthcare entities and support a healthcare scenario, like a patient in deposition, that kind of thing.

So pick that up and took that solution stack forward to establish a test scheduling mechanism for people with the Power Portal in play.

Neil Benson: [00:13:12] Okay. So I'm interested to go back to one of your previous comments about the resource scheduling capability it's built in there. It was part of Field Service originally to me that seems like a pretty close fit to this kind of challenge. But I really love your focus on keeping it as simple as possible.

You've got to scale this thing super quickly. The usability needs have got to be extremely straightforward for an average citizen to be able to schedule their appointment.  Tell me a little bit more about that decision, about why not use some of the inbuilt scheduling functionality. It sounds like you built a slightly custom scheduling feature on top of the Healthcare Accelerator and Power Portals. 

Mohammad Mohsin Khalid: [00:13:52] Absolutely. Yep. Yes. That is what we did. And I think a bit of the decision was surely around the training aspect of the administrators that actually have to interact and create test appointments. Right. And that have never interacted with the Dynamics application or configured one and how to put out resources.

That resource scheduling piece had that training issue, but it was also very bulky with regards to features, right? Like you did not need to schedule a worker to do XYZ and manage resources. The only resource we actually had to manage in this aspect was the test itself. If you see it that way, the criteria or the restriction or the limitation of resources was on the product. Not the resource itself. 

So that's where we kind of, and very interestingly that sort of how we approached the idea of providing the administrators of the state, the ability to schedule just enough appointments or make just enough appointments. With regards to the kits that they would have, and they were expecting, actually made them take it live on Monday night.  There was a go/no go and they were going to see it for the first time in the go/no go. 

Neil Benson: [00:15:27] Wow. 

Mohammad Mohsin Khalid: [00:15:29] Yeah. 

Neil Benson: [00:15:31] That's high stakes stuff. It really is.

Mohammad Mohsin Khalid: [00:15:32] Very high stakes stuff.

Neil Benson: [00:15:34] The first version that you put into production on that Monday night, I presume there's, you're still making changes in production after that wasn't the final version. Was it? 

Mohammad Mohsin Khalid: [00:15:43] Yeah, so we made changes. I wouldn't say a lot of changes, but we did make more changes around just making it the portal a bit more crisp, user-friendly, those sort of changes. But the underlying framework that remained the same, the actual algorithm for scheduling and creating those appointment slots remained the same.

What we did push out five or six releases after that. 

Neil Benson: [00:16:09] I'm fascinated, by why   , trying to coordinate a project to go that fast with 30 people on board, maybe two or three people designing the application and 27 people to figure out the licensing. What were they all doing? 

Mohammad Mohsin Khalid: [00:16:23] In fact it was actually the design, build, test, and UAT were going on continuously. 

 We had a whole Teams chat going that was on fire, I would say. I have never done something so quickly, but it has actually been a very strong team effort as well.  

The whole team being invested in the goal while I think the key part of success for it. Cause, I'm just telling you right now, and also listening to myself, it does sound like a recipe for disaster. Right? But it all worked out because we all had that goal in mind to make someone's life easier for scheduling.

And it was the need of the time. We like just to give you an overview of how it was going on. We had like two architects, me and one from the solution side Richard Clark, Dick Clark . He was part of the team as well, designing and releasing actual requirement level features for development teams to be building them.

And then another set of functional team working with the main administration level of the customer. Reviewing them, and then in a separate team actually stress testing the application with regards to security and scalability, because we can't forget doing all of this functionality we have to make it scalable for a million people to be using. 

 Licensing was my big, bigger concern and yeah, Microsoft was involved. They were champs in it by supporting it. And then just the backend [Azure SQL] DTU servers were upgraded for their production. And before that, I didn't even get to know so much about what Microsoft does for a production environment.

But yeah, we had a very strong team with us. With regards to Power CAT as well, reviewing the code and also making sure that the licensing or any API throttling limits, because they have these service protection and the entitlement API limits, involved at every call. Now they don't come in the way for someone's appointment.

Neil Benson: [00:18:42] So some people from Microsoft dragged into the office over the weekend to make sure that the licensing and everything else is put in place and the scalability, right? We are so fortunate these days to be able to rely on the Microsoft cloud to take care of scalability. But I imagine that scalability is normally something Microsoft predicts over months of adoption and usage, so they can see users slowly joining an environment, beginning to consume an application. Here you've got minutes or hours to scale up and you have to rely on a few phone calls to Microsoft rather than just their normal telemetry. 

Mohammad Mohsin Khalid: [00:19:19] Yeah, absolutely. Because the normal pattern is definitely the resource scaling happens as people use application, and it goes both ways. You know, it, they scale it up or also scale it down. If not a lot of people are actually using the environment, but in this case we had who we didn't have that time.

So it had to support million plus day one. 

Neil Benson: [00:19:43] Were there any wrinkles in there? Did you hit any challenges in terms of scalability or performance that were noticeable to the citizens in the state? 

 Mohammad Mohsin Khalid: [00:19:50] No, not after go live. Previous, before go live, I think, on Sunday we ran into some scalability issues and that's when we had to get the server itself upgraded to a higher end DTU .

 Neil Benson: [00:20:02] That's pretty impressive though, that you've managed to hit production in a couple of days and have no wrinkles since then. Well done. 

Mohammad Mohsin Khalid: [00:20:09] Happy we could help people 

Neil Benson: [00:20:10] Using Microsoft portals was there an authentication requirement here? Were these users authenticating with a Microsoft account in order to identify themselves or were they creating an account during the process of registering for an appointment? 

Mohammad Mohsin Khalid: [00:20:23] They were considered anonymous users and they were creating their account, like registering, with just entering their contact information and then scheduling. So they were not really authenticating or becoming an external or internal user of the portal. 

 Neil Benson: [00:20:40] Does that not present a challenge then if they need to come back and modify or cancel an appointment? Could they do that if they're not being authenticated? 

Mohammad Mohsin Khalid: [00:20:49] We had I, I don't think I can talk a lot about it because it's still in play, but we had a GUID-based identification mechanism designed within the texts and notification links that they would receive of their appointments.

Neil Benson: [00:21:04] Right, okay. I've seen that approach used. For example , a really simple one is whenever you go to unsubscribe from an email list, sometimes you click on the unsubscribe and they say, "Oh, please log in to amend your subscription preferences". You're like, "Oh, what!?" Other times, the link that they give you to click on, it's got an identifier embedded within it and they automatically know who you are and what the action is you want to take, which is to unsubscribe it, and it's done. So It sounds like you were taking a similar kind of approach. 

Mohammad Mohsin Khalid: [00:21:32] Yep. You could say that.

Neil Benson: [00:21:34] The Healthcare Accelerator, I took a look at it for it, a healthcare organization here in Australia. It seemed to me a little bit US-centric, which is understandable. There's a lot in there about insurance and payers and stuff that only exists in America. What benefits did it provide in this scenario? What kind of building blocks do you get from that Healthcare Accelerator that means you didn't have to build something yourself? 

Mohammad Mohsin Khalid: [00:21:56] Fortunately, I was part of the development team of the Healthcare Accelerator. 

 I was working at Mazik, the known healthcare partner at the time, and Microsoft approached us to actually create the Healthcare Accelerator. Like, Microsoft does ask a lot of ISVs, development partners, to create accelerator solutions.

So I was part of the development team that actually built the Healthcare Accelerator. I knew a lot of insides of it, as well as I knew how lightweight just the data structure model could be helpful in this case. The Healthcare Accelerator itself doesn't really do business logic. Right. What it does really well is it provides you a FHIR- based data structure for healthcare entities.

For those who are not familiar with FHIR [Fast Healthcare Interoperability Resources], it's the healthcare standard for exchanging data and keeping healthcare data, which is FHIR data, patient healthcare information into digital format. So that's like the defined and agreed upon standard for a data structure. 

So Healthcare Accelerator follows that. And if you have an application using that standard, the moment you want to interact with any other system, in the U.S. at least, that is following the health care standards. You're going to not have a bigger problem, because you're going to be using the same kind of terminology, same sort of way to keep your data structured.

So that's what the Healthcare Accelerator provided.

Neil Benson: [00:23:32] There's no requirement for Dynamics365 application underneath? You can just deploy that straight on top of a, I don't know what the right word is, a naked Dataverse environment? 

Mohammad Mohsin Khalid: [00:23:42] Great question. So the first release of Healthcare Accelerator did have a requirement on almost all the Dynamics applications. But it was stripped down as Microsoft is promoting more the Power Platform and the Dataverse piece. So it was stripped down to just run off of Dataverse and that's what made it a perfect fit for, you know, a lean and clean system.

Neil Benson: [00:24:10] I've looked at some of the accelerators and some of them have two versions, one where you need to take a dependency on Dynamics 365 Sales or Service, and then another one, which you can install on a clean Dataverse environment. Yeah. It's a great option. It really must make it more difficult to maintain, but it's a great option that Microsoft has provided the ecosystem. 

What was the reaction like from the state's I.T. team on Monday whenever they saw the application for the first time. 

Mohammad Mohsin Khalid: [00:24:41] Their I.T. team was pretty well-versed in Azure. Right. So I think they definitely loved the fact that they could just use Azure security groups to share the canvas applications, because their main worry was, how would they share the canvas application or have people access those applications who are on ground at the site. Because mainly those from our end or the state's end the people who are going to be the heavy consumers of the application would be the ones who are checking people in while they're sitting in the car with a QR code. So we provided this QR code functionality where a canvas application is scanning the QR code and the appointment is being identified. 

The I.T. team took it pretty well. So we did run them through the deployment and do it for them for the first release.

There weren't really a lot of challenges there, but I think their main concern was solved by how Azure does interact with Microsoft Business Applications. And they're not like two separate worlds and the identity is the same and stuff. So that was good news for them.

Neil Benson: [00:25:48] The citizen turning up for the appointment is given a QR code that they display on their device. And then the healthcare worker has got a QR code scanner, and they can identify the patient and the appointment in their car whenever they turn up the test center. Is that right? 

Mohammad Mohsin Khalid: [00:26:05] Yeah. So the canvas application, the base application, they were running it. Each healthcare worker they did not, the state, did not have funds or the resources to have custom devices. So everyone's own mobile application. We just provided a guidance to install the Power App and open up the Power App on their phone and use their phone's camera to scan the QR code while in the application. And then the application would read data point and then bring the person's information in for them to check them. 

Neil Benson: [00:26:37] Cool. Isn't that amazing? That, so they, the Power Apps player on works on almost every mobile device and Microsoft's tested a lot of them, but you don't have to worry as an application developer, you don't have to worry about screen sizes or resolutions or form factors. You got to trust that it's going to work on 95% of these healthcare workers' devices over which you've got no control. You don't know which manufacturer or operating system it's running on. 

Mohammad Mohsin Khalid: [00:27:03] Yeah, absolutely. And just that it works on the browser as well. So a lot of the workers, they don't even install the Power App on the app store. They just use it on the browser. 

Neil Benson: [00:27:15] Okay. And you can still get access to the camera and things through? All right. I didn't know that I thought to get access to the hardware you had to use an installed application. Cool. No Windows Phone users or Blackberry users or anything you ran into?

Mohammad Mohsin Khalid: [00:27:29] No, not that I'm aware of.  We had minimal challenges with the canvas app. That went very smoothly. The maximum challenge was we had was with some users who  did not really want to use a digital device to check people in. They brought everyone in, whoever could contribute to actually carry this humongous activity out, for the whole state.

So not everyone we should realize that not everyone is digital. You know, are using phones to interact with all sorts of scenarios. So that's where we had some challenges or the state had some challenges, I think, to bring in those people on board, give them devices, give them a few days of training that, yep, you have to do it digitally. You can't just write their name on a paper and pass it on. 

Neil Benson: [00:28:24] No, it's not gonna work. It's definitely not going to work at that kind of scale. So once the patients come in and they've had their test, the test results are not coming back into this Power App environment. I presume you're sending the patient's contact information through to a different application that the state has for conducting the testing and getting the results back to the citizen. Is that right? 

Mohammad Mohsin Khalid: [00:28:45] Yeah. So in the first go live we did not have any relationship with what happens after the test is done because once the test is done, they have a different mechanism to send it to the lab. But in the second release, we provided an integration with a blockchain identity provider. Which actually enabled them to validate identity and then receive test results back to the same identity through a blockchain verification mechanism.

It was, it's a pretty kind of complex scenario, but on the very high level of it, we would send a, once the test is done, we would say information to the blockchain identity provider. And then, from then on that identity provider would interact with the patient and then send them the results from the lab.

Neil Benson: [00:29:36] Mohsin you are the architect who likes to keep things simple.  Had you done a blockchain integration before? It's just something you'd got experience in or did it just seem like a nice, simple way of handling the verification of patient data? 

Mohammad Mohsin Khalid: [00:29:49] So first it was the state spec, but secondly, the integration part, it was a very simple REST-based API service for us to call. So no complexity added on the solution itself. And then the complex part was within their own identity verification mechanism, which they're pretty famous for doing so the state trusted them and they carried out that whole scenario for people.

But yeah, not no integration complexity there.

Neil Benson: [00:30:20] So  what lessons can you learn from this kind of project that you're going to be taking into future projects? 

Mohammad Mohsin Khalid: [00:30:25] I think the biggest lesson right off the bat that I learned is that now a lot more is possible with regards to the speed of delivery. And we don't necessarily have to think of digital transformation as a five to six months, or a year to two year, activity to take on and then reap the rewards of, right? With Power Platform, a lot of stuff not just is provided to you by Microsoft, but a lot of stuff within the platform is available to adopt quickly and then change or develop over. So the full solution layering and stacking mechanism, Microsoft has given. That you can adopt Dataverse by itself as well. You can take on applications, not with regards to just licensing piece of it, but also the solutioning piece of it, that you can take on applications separately. Like first, for example, in the, in this case, I'm saying you can just adopt Customer Service, then goal for the Sales module, then go for Field Service previously, you would get a customer engagement license , you will be paying for the whole thing. 

So those sort of barriers and a questions have been answered. All in all, what this is combining to do is enabling rapid data transformation for people. So for this test scheduling, we went with a solution that worked in five days, but for an organization's own business transformation and getting on the cloud, they do not have to plan very far off. They can reap the rewards quickly, month to month, and start with MVP because the technology stack makes it possible. 

Neil Benson: [00:32:20] You're not going to be delivering all of your future projects in 48 hours. Are you? 

Mohammad Mohsin Khalid: [00:32:25] Hopefully not, but yeah, if you need one I'm the guy! 

Neil Benson: [00:32:30] Yeah, it sounds like it. Well, and you got some recognition for this project as well. I believe that you recently received the Microsoft FastTrack Solution Architect award. Is that right? 

Mohammad Mohsin Khalid: [00:32:39] Yeah, yes. I was privileged enough to work on this project and because I worked with the Microsoft Power CAT team heavily, they honored me with this award and it is for the Power Platform itself. 

Neil Benson: [00:32:52] Congratulations. 

Mohammad Mohsin Khalid: [00:32:53] Yup. Thank you.

Neil Benson: [00:32:55] Working with the Power CAT team, is that a team you're working with alongside FastTrack in this project or is it just Power CAT in this case? Because I think FastTrack is normally just on Dynamics 365 projects rather than Power Platform. Is that right? 

Mohammad Mohsin Khalid: [00:33:09] Yeah, so I think there's some changes going on within Microsoft and those lines are getting blurry as we go on. Right. Not one area or department or team was involved in this. Power CAT was involved, the Microsoft product team was involved and FastTrack did come into play for a few things, because this was just customer success all across. Customer success was everyone's goal.

Neil Benson: [00:33:33] Well, it's an amazing feat just to assemble a team of 30 people in a space of a few hours on a Friday afternoon. To find out who's available? Who's got capacity? Who's got the skills? Who's ready to spend all weekend on coffee with no sleep? Yeah, it's... 

Mohammad Mohsin Khalid: [00:33:47] Yeah, no, we didn't keep everyone up, Neil. Maybe we're not that it's not that hard. Not all 30 people went up the whole weekend.

Neil Benson: [00:33:55] And in the future releases you talked about second releases and so on. Was that only a few days as well? Or did you have a few weeks to improve the application after that initial deployment?

Mohammad Mohsin Khalid: [00:34:04] We didn't have, I think more than a week gap in a release, like every release was within a week, because yeah, because it had to support all the oncoming requests by the state, as well as  customer feedback, user feedback, or if there was any issues in the application.

Neil Benson: [00:34:26] One of the challenges you must have had when you start a project that quickly is you don't even get time to consider what your approach is going to be. Communicate that approach. Maybe train everybody on it. Make sure Azure DevOps is in place and managing requirements and get your ALM story straight.

You're probably just diving straight into it. You're just working on instinct if you're having to start a project that quickly looking back on it now, any really helpful tips or alternatively, anything that didn't go so well that you would fix in future? 

Mohammad Mohsin Khalid: [00:35:01] Absolutely. So I did have a few lessons that I take forward still. One key area that I see that was kind of the success factor in this, like you mentioned, we definitely did not have time for DevOps, ALM or reviewing a lot of the requirements coming in. So the key success factor again was the simplistic approach to keep it very lean and very simple and make an MVP, an actual MVP. Right? So not have any functionality sitting out there that's not going to be used. Only have functionality there that is going to be used. So it was that made it easy for us to keep it manageable. Right. And that's something I take followed from it as well, because a lot of times I've worked in the, I don't know if, you know, I worked in the ERP side of things as well for Dynamics for seven years, I've done that. And a lot of times we are over selling or over deploying the application as well. And I think the Gartner statistic as well, like usually 30% of functionality is used by a customer in an right? Correct me if I'm wrong.

Neil Benson: [00:36:23] I thought you were going to say 30% is never used. You think only 30% is used?

Mohammad Mohsin Khalid: [00:36:30] Absolutely. So what Dynamics applications even delivered there's a ton of features that no one even touches or goes there. Right. But they're deployed, they're sitting in the solution. They have one or two features being actually used from multiple applications. So that I think that approach is not going to work as we go for more rapid deployments.

And customers also are now realizing that the timelines are reduced. They do want to see quick results. So that sort of approach is not going to work for when you're trying to deliver an MVP. So you have to track carefully on the architecture side. And know what you're deploying and exactly what is going to be used.

And then I think the other side of the lessons learned would definitely be that in order to do something like this specifically, right? You need the right people available, like you were mentioning. So we fortunately we had Power CAT available on the line, on call, and we had a pool of resources that were very passionate about this and volunteered their weekend to come in and help out.

But without that team, you can't really do this sort of activity. So the team you will have to count on and on for success in a project.

Neil Benson: [00:37:55] I've got a kind of a philosophical question for you.  The United States has got 50 states and districts. All had a similar requirement at around the same time to build a similar kind of application. In Australia, we have seven states and territories and I've seen time and again the same states have got overlapping requirements for an application, whether it's for managing COVID appointments or court appointments or real estate transactions or something that just gets repeated over and over. 

Mohammad Mohsin Khalid: [00:38:25] yeah.

Neil Benson: [00:38:25] I'm wondering, do you think this would have worked if it had been a federal project and you were going to roll out an application across all 50 states and districts of the U.S. or did it only work because it was limited to a single state? 

Mohammad Mohsin Khalid: [00:38:39] It actually did roll out to other states after they saw the success because they were in trouble with everyone was trying to do their own thing and try to use what they have or build something in-house, but not everyone succeeded.

And one of the U.S. states failed really bad with another solution that was also on Power Platform. But it had actually used the USR approach. So not digging the name of the state, but yet they did have a failure on the same kind of platform, but just the approach maybe made a big difference.

And then we did end up deploying it to multiple states and to others countries as well, to cities and other countries as well, the county. It did work for multiple scenarios. Because, again, I bring it back to the same point because it was very lean with just the absolute, basic process that was there. Any flavor that someone wanted here and there could be easily added.

Neil Benson: [00:39:40] That's fantastic. Well, congratulations on its rapid adoption worldwide. Mohsin, thanks very much for joining us and sharing your story. Is there anything else you'd want to mention and how can people follow you and find out more about your work, if they'd like to do that?

Mohammad Mohsin Khalid: [00:39:55] Yeah, thank you for having me, Neil. So people can follow me on Twitter. MohsinKhalidAX is my handle. Or other than that, they can come to the Imperium Dynamics, website and review what sort of services we have and the team I've built around me. And we're definitely focusing on this same sort of approach, building on the Power Platform with the right skill in place to deliver digital transformation in a rapid manner.

Neil Benson: [00:40:25] I wish you every success with your new venture. Good luck with that. And thanks again for joining us. 

Mohammad Mohsin Khalid: [00:40:30] Thank you, Neil. It was great to be here.

 Neil Benson: [00:40:35] Thanks Mohsin for sharing that story about the COVID testing appointment booking portal that you built for one of the U.S. states.

I booked my own COVID vaccination appointment with Queensland Health yesterday. So Queensland is my local state here in Australia. Guess what? They're using Power Portals too, perhaps that's no surprise to you. In fact, we probably shouldn't be surprised if their design pattern was similar to the one that you designed for your state government in the US.

We've got lots more great interviews lined up with more Microsoft customers and partners, building amazing applications. And if you know someone who you think should be telling their story and sharing it with us, send me a LinkedIn message and let me know, or have them visit customery.com/guest. 

You'll find show notes for this episode at customery.com/036. 

Please remember to follow the Amazing Apps podcast in your podcast player so you don't miss those episodes we've got coming up. 

I really appreciate you letting me into your ears. And I hope you find value ideas and inspiration in every episode. Thanks for listening until next time. Keep sprinting.