Software Development 101 with David Burkett from Working Mouse Ep006

[Music] hello and welcome to Red's Business and Technology podcast I'm your host Jackson Barnes and I'm your co-host Brad Ferris and today we sat down with David Burkett who's the chief growth officer from working Mouse speaking everything in software development in this episode we cover off what to do when considering custom software development what not to do when considering custom software development over to you Brad what did you what are some key takeaways you got from the chat yeah I think it was really good I'd never met um Dave before so it was really good open conversation I thought he was really open with his advice and probably the key takeaways I got out of that is how important it is to define the problem you are trying to solve before you go down uh go down the path on a software project um once you've defined that problem also how important it is to then get the business case right before you go forward because you can really burn a lot of cash really quickly if you don't do kind of the proper planning and I think that's something that we instill in all of our projects don't shortcut things because if you do it can quite often end up causing you more in the long run so I thought it was really good I really enjoyed meeting him and um um potentially working with him in the future there's a good takeaways I think the um has to be a real business case for it that's something I really really took away let's jump over yeah thanks David for coming in really really appreciate it um if we can start with maybe introduction to your background before we jump into working Mouse it's probably a good place to start yeah fantastic thanks Jackson and thanks Brad and uh thanks to Brad for having me along really appreciate being here uh so introduction to my background um so firstly I'm not a software developer um I would gauge myself as someone that knows enough to be dangerous and I've got a lot of project experience so my background before working mouse was working in b2c for Apple uh previously also did my own software development technology startup back in about 2012. that involved a trip to China Chinese software developers illegally mapping applications in China which is a whole other story um and a massive failure of an application that we launched um and a lot of learnings wow yeah learn what not to do yes exactly and then you uh joined uh working Mouse yeah brought those learnings in yeah nice so do you want to touch on working Mouse um we had a little pre-amble around 70 people in the organization software development but you want to tell the working mouse story a prison-based success story like red just around the corner a couple of suburbs but how did it start and uh when maybe the foundations and where it is now yeah lovely lovely uh so working Mouse started out of the founder my business partner Dr Urban ascot's PhD so Evan was actually working back in the UK approximately about would have been 12 15 years ago Contracting as a developer for uh Master he's a heap of big projects coming into the office and doing the same thing every day and getting paid an extraordinary amount to do that as good as that was he thought there's got to be a better way to do this so Edmund previously had a masters in artificial intelligence and machine learning from qut so he came back to Australia and went to uq and did his PhD at uq in model driven software engineering out of that Evan's PhD came out uh MD SAS um so model driven software engineering as a service um and uh he went and pitched that to a number of investors and they said we've got no idea what you're talking about mate yeah I need to go slow on that one actually Dave I didn't keep up then I'd be Brad and uh they said go go start a business so Evan started working now said use the technology that you're talking about to start building some projects and some products and take it from there so that's how working Mouse started when I and the name the name or that are we coming to that yeah or the names actually um was working mouse was a previous Incarnation within Brisbane it was a previously a development agency before that so um he actually purchased the ptyt from his brother who owned at the time and that's how we got the name working just yeah it was just about her convenience there's a PTY we'll pick that up yeah it's boring Dave you should make a big fancy story about it yeah that's because we do that five thousand clicks a minute I don't um so um yeah so Evan started the business up um and started working on projects um and uh we we they got to about three three or four different customers then they suddenly lost two or three um when the first GFC hit um and went right we're a bunch of nerds sitting in office we don't have anyone kind of out there spooking the business and which is uh the point that I became involved in the business and now I kind of head up the growth team so since then for working Mouse we've grown to yeah approximately 70 and 10 delivery teams we've been quite lucky we've had some good contracts so we've built the non-materiel and now material procurement platforms for the Department of Defense and we're now working with customers like Hitachi on their automated haulage system for their automated vehicles uh bluescope steel and a number of other kind of large clients in regards to the technology and the PhD we have kept on developing that and we realized about two years in that we were building a service business and not a technology business so we decided at that point we were in a very conscious decision to split the technology business and the the services business apart so at that point we founded codebots which is our technology service business so we actually run two businesses within the group we've got the kbox technology which we've been r d for about 10 years now and is getting fairly sophisticated and then we've also got working Mouse which has kind of been going for a strength to strength on project so working Mouse is services and cobots is IP effectively yeah that's it yeah kobots is I would describe it as a software development platform as a service for Developers right and do you are you selling that yeah yeah cool yeah do you white label it or it's codebots save the product in the market or so how would I describe the product yeah so codebots is yeah a model driven engineering platform as a service and what that enables us to do is to effectively train a bot on something that we write by hand so whenever we write a piece of code we can go right well we're going to reuse this piece of code in the future on other projects within this Tech stack um if the case is yes then we go right let's train the bot on that piece of code and then we publish that to the web IDE so that can get reused from a model so the point behind model driven engineering is that we bring all artifacts up to be first class artifacts so usually with the software development you come into a room you whiteboard it out and then you write down the requirements you go yeah that sounds really great there's our scope there's our spec we've got our requirements backlog quickly clickable prototype and uh yeah let's go off and build that and then that kind of gets left by the way so I hadn't forgotten about as the project evolves and iterates the idea is that the model is kept up to date so the model has traditionally when we're talking about a a bot model with a software development bot we have a entity diagram which is like the attributes the relationships on the server and we can add on different behaviors that we've trained security for Access and then we have a user interface diagram much like a traditional low code development platform but the difference is that we don't obfuscate any of the code and what that means is usually when you use power apps or you use something like that to kind of create an application you don't have that control down to the code they control that they guard that and the developers go oh I can get you this far but you've got to follow that process here because they've decided to do it this way the difference here is the developers get the code and the bot works like another developer and commits the code to the git repository so you can go in there and tweak it all you want go and tweak the code and Away you go so you're going to have that control code level at the platform level and also within the target application that it writes yeah it's good this means you if you get to do something in powerapps you're a bit limited to what you eat how far you can go whereas from what you're doing fairly unlimited exactly yeah so like powerapps workflows right you come and go cool I can follow their workflows here uh to this point and you go well it doesn't quite do what I wanted to do okay we'll just go into the code what percentage of the business is across like the codebot versus the other side of the business the majority is still on the services side yep um we're still finding our product Market fit for how the platform works and kind of Licensing Bots if that makes sense yeah that makes a bit of sense yeah so what's working Mouse point of difference in the market from other software development companies yeah probably the cobots so obviously you use it internally yeah definitely yeah yeah it's your secret sauce secret Source day-to-day that's what we use yeah the point of difference is that we can use that so we are up to what I would call our third generation Bots with codebots so we've got their gen Bots we've got two bots on there that are public bots so uh one is we call C sharpot and you guys will be quite familiar with I'm sure in regards to that's Microsoft's open source technology stack and then we also have a Java Springbok which springbot which is very similar to Google's Tech stack so it's what kind of a mix of what Facebook Google and Microsoft have open sourced and we use those we train the bot on those and we can also license them out as well right yeah it gets us a fair bit of velocity at the Box yeah that makes sense so what's the process of getting a piece of software developed so we've got one of our clients or someone approaches you and they're like oh hey I need something to go and do blah where do they start what I'm going with this is one of the most value for listeners of our audience which is a fairly General across technology how do you start that process yeah definitely it's a great question um so if you're starting from I have a problem that is the best place to go from rather than I I know what the solution is going to be so if we talk about software development we usually these days talking about product development which is where you're bringing a piece of Ip to Market as software so in that case what we do is we start with the problem we all agree on what the problem is is part of a problem statement and then we go into Skype now there's lots of different ways to skin a cat when it comes to software development in uh you know waterfall project approach or agile project approach and you've got different flavors of agile but the thing we need to do first is figure out what the solution to that problem is so we go into a scoping process at that point and we have not only the developers but also designers working alongside you to figure out what the solution to that problem is so the way that we do that is we go for a discovery process so Discovery we go out and we interview different users we observe what they're doing and we go right okay well we believe that this is going to be the correct solution we prototype that out we test it and then we go here all right that's going to be the best thing to solve this problem and then at that part of the back half of scoping we get all of the things ready so we we have a way of working and the way of working says you can't go into development until the artifacts are ready and they meet the definition of ready so the key artifacts as I kind of alluded to at the start is backlog of requirements the backlog of the requirements is the key one that the developers use throughout the project and that's basically user stories where as a user I want to do this to achieve this within the you with the backlog for the user stories you then have the implementation notes so the implementation notes for the developers to go oh yeah that's how we're going to implement this but the key thing and this is the big thing for customers is to understand that user acceptance testing and user acceptance criteria so user acceptance criteria you do that at the start when you're scoping yeah we have that all Skyped out and user acceptance criteria enables the customer to then tick off when the developers have done it correctly and if you don't have that that's where you can fall into some kind of big problems so there's a lot to unpack in that in that little bit there so first of all the so your Discovery process up up front do you do that as a kind of as a one-off separate engagement to go like cool like you've come to us you've got a problem you want to solve you might not have any clue how much is it going to cost or how much efforts involved on both sides yeah and will you do that as a kind of cool well we'll we'll spend a week or however long we'll do Discovery you'll require this much time we price that separately after that we'll have a bit more information to go into um the discussion around okay how much work is this going to be how long is it what's going to cost the Technologies all that kind of stuff is that how it works yeah definitely um so it's it's separate from development obviously but the key thing is having the same team do the scope that's going to do the built if you have a different team doing the scope and then they hand that over to a new team you as the customer have explained your domain that you're the expert in and then you have to re-explain that again as to why those decisions were made during scope so it's beneficial to have the same team scoping it and not just a designer you've got to make sure that you've got the developers in there so they can go and de-risk and Tech Spike the different elements of that the the way I like to explain software development especially these days whether you're making a web app or you're making a mobile app or something like that is you don't have control of the Technologies and this is what makes it so complex it's like you're building a bridge and you don't have control of the land on either side of where you're trying to build the bridge between and you're trying to engineer it as perfectly as possible and they change the height over there they release an update over here things are moving so fast yes there's too many unknown notes yep yeah it never it's literally never ending software exactly yeah it's an iterative process right software is never done you you build it to a point to solve a problem and you can keep building you can keep building you can keep building yeah minimal what do they say minimum using like using minimal viable products yeah there's loads of different things you can say but yeah that's generally the approach you do is get to the minimum model product and and then enhances like a phase two three after that yeah exactly so the process is that we'll we'll Escape out the build one or the MVP um we'll then estimate that we'll agree upon that estimation and then we proceed into development for that time and at the end release to production and then we move it over to our devops team and devops the devops guys Provide support so if you want to make enhancements or you want to Lodge bugs and things like that and get it resolved you can go ahead and do that via like a level two service desk so are you guys um are you kind of full stack service in that regard like you'll do the design the ux UI yeah devops Dev exactly yeah completely cross-functional software development and the Consultants to do the Discovery work all that yeah exactly yeah yeah and we get the developers to do the discovery work alongside the designers because they're the ones looking out to build it yeah yeah I want to get a bit more into like the the elephant in the room here around pricing because I feel like it's really common maybe misconception or maybe just how it is around people going oh yeah we need this but we don't want to go to a software developer because it would have cost so much ridiculous amount of money to get this stuff built and you no doubt would get that um I guess challenged all the time when you first had that initial conversation and you're trying to qualify whether it's going to be worth either you know have like the budget to get this across the line how do you I guess combat that and is that something that you get um like that scenario happens a lot yeah definitely um so the first thing is there needs to be a good business case the business case to build a product or a custom piece of software needs to be very strong and it's it's often hard to find that strong business case um so if when I say the business case and I mean why would the business case not be there the business case would probably not be there if you don't actually need to own the IP and develop your own product so for example if you can find something off the shelf like a cot solution it's a custom off-the-shelf solution and then utilize that or white label that and you're only using it to increase your sorry lower your costs because it's saving you a certain amount of time why would you need to develop that as a piece of software if you're not going to license it to other people so look in the market first see what there is and then if you can't find something and the business case is strong come with us so I've got a good example of that we built if you guys ever been to Modern Island and gone on the mycap ferry yeah so it's the 4x4 Ferry goes out from border Brisbane so Liz came to us about six years ago and she said Dave there's there's no one in Market that'll enable me to build a ferry booking system that has variable pricing or dynamic pricing so they can price based on demand and we said sure we can build you a ferry booking system and we can do that based on demand that's absolutely fine because there was no one else in the market that could do that we built the product and obviously software development when it comes to cost has a really large capex it's a significant amount to invest up front which is why there needs to be a strong business case but once you've got that you're not then paying the license fees per seat or anything like that so the use case with lids was that yes there was a large capex but once you'd appreciated the software the return she made was 10 volt so she not only made a return on the fact that she was able to optimize her Revenue through Dynamic pricing but on top of that as well because we had the booking platform they did they moved from going from managing their business via telephone calls in a service number to everything being on the booking platform so that saved to about 200k in salaries per annum all right so you do a full Roi on before you even engage past the initial scope you would do an Roi case back to the client of go moving forward we would have that full discussion and we'd do as much investigation as we can to make sure it's worth actually proceeding forwards so yeah I was gonna so in the scope of your service when the client comes to you like in the next example and she says I can't find anything I don't know will your team actually as part of Discovery or whatever you want to call that initial engagement actually go out to Market and go well actually there's something here you've paid us x amount and we've done Discovery but actually you should probably look over here exactly yes that's it and then yeah then why would why would we build something for someone if there's something out there that's going to solve the problem already for them yeah I would much rather be transparent about that and be invested with a serious business case that makes sense and so what are the pricing ongoing so say there's a large investment you said for building something completely custom what about ongoing is there is there like fees to maintain and Patch those applications or how does that work yeah definitely so like as we said before software's never done so there's a couple of different fees that you need to take into account so firstly you've got your hosting fee so Cloud hosting these days we're agnostic we don't care where you deploy it you can deploy it on Prem if you've got the capability or you can deploy it in a managed environment such as Azure or AWS or if your it service provider has that sophistication they can manage the deployment for you as well right so we usually say sign up to a cloud provider and we do the deployment we manage that we can manage the deployment through the devops guys there and then you've just got to pay your usage fee to Microsoft Azure or Amazon AWS so that's that kind of Hosting on top of that as well you need a retainer in there for the support desk and then access so the big question is well how much if is software going to cost me on a per random basis to kind of support and we usually say the best thing to do is to take a budget of the build price between five to ten percent of the build price per annum and factor that in for support and enhancements so if you've got that budget there you're able to always go yep okay we know it's going to be 10 per annum so if it's a 300K build let's allocate 30k per item to do enhancements and bugs on top of that as well software's always changing so there'll be a new release of the open source technology what we'll actually do is we'll within the technology set we use with codebots we'll keep an eye on that and then we'll go right let's update cobots to the latest version of react or something like that which is a client-side language how do you actually how does it so that's how that changes how does your customer know that cost is coming or do you not charge for that are we so what we do is we update the bottom we wear that and then we go right we've gone from you know react 16 17 to 18 which has just come out and we'll then update that on the bot and then we'll tell the customers hey we've updated our bot to this did you want to update your application to the latest version Yep this is the estimation we'll give them a report as to hey this is the pros and cons of moving to react 17 to 18 this is what you're going to get out of it so it's like these are the security updates and things like that yeah and then it's just the labor for us to go yep update that application to the latest version the bot so the big thing with that is you then don't end up in a situation where you're on a legacy version when it happened in three to five years that is no longer secure and no longer patched so your your technology is always on the latest and greatest and you don't slip behind and we've had customers previously that have had custom builds done for them and it was on angular one and there was no migration Pathway to angular 2 and so on and so forth which is a Google client side language so we had to rebuild the whole product for them play into doors oh wow yeah so that Legacy creep in patching super important as well so I imagine that conversation is also part of that Discovery or or at least maybe just post Discovery when you're waking up all right well what's the technology stack to kind of future proof this to an extent there's always going to be the maintenance in the in the development but I imagine when you're choosing that technology um and I guess where I'm going with that is that probably why it's important to get the right team that can have that conversation up front and not just go down a path and you end up after you've sunk however much money into a project actually well I'm stuck with this thing that I can't do anything with and then I end up doing the project twice because I got to go to someone who knows what they're doing yeah and the biggest thing as well is that because it's changing so frequently there's always a new technology coming out that developers want to play with yeah and it's making sure that you're not going with you you know the new XYZ technology that is really popular today but in two years time no one supports it so the way that we do that and the way we decided our Tech our core technology stack so when someone comes to us and says can you build this we say yeah this is the technology that we use is we looked at adoption amongst developers as a worldwide community so these Technologies they're all open source they all have communities so c-sharp graphql which is the API layer and react as the client-side layout for our c-sharp bot they are the most commonly used Technologies throughout the world so what that means is that you're not going to find lack of developers that can support that it's not some weird obscure language that you're going to go okay this seems great at the time but then you can't get anyone to support that in five years time I do want to touch on actually the uh finding talent because going to 70 people is big but before we do that the you touch I've seen some horror stories of people who get like a one-man band developer or something on the side to go and develop something for them and then one year later they leave or if it's a Director of Business Who develops something they run away there's there's I've heard a lot of horror stories over the years what are some of the mistakes that you've run into um of like what not to do when looking for a software how much time do we have Jackson yeah give me they give me this two minute version like I said it's the wild west right so there's minefields everywhere yeah um it's it's a tricky one some of the core mistakes um are yeah developing solo and they're not having documentation not having documentations probably the biggest one so um it's an inherent problem because everyone just wants to be building especially in software development I want to do the cool stuff I want to build the stuff um what I don't want to do is write the tests for what I built and then write the documentation for what I've built so someone can interpret that and understand that in the future and that's a big one um so yeah okay you can go to a software development agency and they could say yeah we can get this done in X weeks and compared to us that would be significantly less from our estimations and the reason that is is that we don't only estimate to build it we estimate the time to test it and put the automated tests in place and then we estimate the time to actually document it so you're not only getting the application like I said at the start the the other artifacts are just as important as the source code so the documentation the tests are just as important yeah we see that in a small stat scale here on it right where we inherit a network for example and there's no documentation on anything we're like oh I'm going to go and look at full Discovery again of what's going on if it was with a smaller managed service provider or one-man band or something we've run into the same thing but I imagine with software it would be a bit harder to unbake and and learn what other things like have you like top mistakes have you come across um that's probably been the biggest one um and the the other ones um are as I we alluded to before is not patching and modernizing so letting it kind of drift into Legacy where it kind of becomes too hard yeah and they you can't touch it at that point but even before that you know something that I've seen time and time again and it's both for software it's both for what we do is that people don't understand what's involved right so they'll go to working Mouse they'll come to Red then they'll go to whoever software I've seen it several times where people just they'll get an offshore quote right and then they go whoa whoa well hang on I'll just use some numbers hundred dollars over here twenty dollars over here like what's the difference and the difference is all the things you just described um yeah it's quality the quality of the documentation the quality of the testing the automated testing um things like that so I actually find that quite a big problem is trying to articulate that quality to someone who doesn't really understand the process and getting that across but I've seen a handful a few yeah of software projects that they end up yep they go with the 20 option and just get down to pass so they they lose out because they don't get the product to Market somebody might also come they lose out because what they get actually doesn't work anyway and if it's an established brand it could actually hurt your brand because you've put out a dud and what you end up doing anyway is just blowing that 20 grand and going to have to spend the 100 anyway or whatever the numbers are I'm getting it redone properly so um I have seen that many many times unfortunately yeah that's it the the complication when you're going offshore um is that the communication the culture and then you have to kind of work asynchronously because they're usually in a different time zone and um you have no idea and control over quality yeah um and you can't come sit in a meeting room whiteboard something out ensure they understand that's a big thing but that has a place though like offshoring um a mobile app development when it's very high risk and you want to just test whether that's going to be a market fit there um go for it makes sense you can rush to Market with something do a perfect concept with that and then maybe you Circle back to an organization like yeah he misses it right yeah but it's all based on the use case right if you you go and build something overseas um you can't then expect to license that into a government Department in Australia because it's not going to match the quality standards you know like I think at least it needs to be oh there's some examples and it depends on the technical capability of the business that wants to get something done but that hybrid model if you're really going for for cost and again probably for the prototyping or a small project but I guess the message I'm trying to say is like you kind of get what you pay for totally and um just to be aware for listeners that you know there's time there's a smart people that work on these things and it is actually quite complicated behind the scenes especially to get it right because it's very abstract there's no black and white in this stuff yes exactly I wanted to touch on um a cyber security side because we're big on cyber security at Red yeah it's one of the core functions we do and with software development you know you're putting your applications on me different environments hosted wherever how do and I imagine or two-pronged question first you probably have you had like a increase recently around people asking you cyber security questions with everything going on in the market and then B how do how do you work out as working Mouse combat those concerns people might yeah definitely um so we've definitely seen an increase from government so government very very restrictive in regards to your qualifications so the big one that the government wants to see is ISO 27001 and they want to see that check they want to know you've been accredited and we're on that Journey we're not there yet there's not many people that have completed that journey in Australia so it's a hard one to keep up with unless you're kind of a tier one provider you're a Fujitsu or someone like that so that's the the first one from from that perspective so how we get around that is we partner with the tier one provider so we've got a number of different tier one provider Partnerships right so if you're a government agency and you want to use a local SME like working Mouse to develop your product or your software we can partner up between someone like Fujitsu and ourselves they can provide the iso 27001 product management and we can then do the development work that's cool in regards to our private clients how do our private clients know that we're meeting security standards and there is a lot to it as well not only on kind of the the database level hosting this the application um and we're very transparent about that there is a way in which you can get things tested fairly easily so there's a number of different providers around Brisbane and Australia that will do third party penetration testing and we've had a number of our clients go to their get it penetration tested to make sure that what we're creating for them is nice and secure and the deployment is secure they can't inject any SQL or anything like that and we're meeting those standards how many of your clients get penetration testing done like it's like half or all or I'd say probably probably four in the last two years yeah okay um so not it's not a huge amount there's a lot of trust there in the relationship they know that we're at equality in terms of standards that we follow to keep up with security as well so we develop to the O wasp top 10 which is basically the top 10 things you can do to protect software and if you uh following the AOS top 10 you're pretty much covered so what was that that's a framework yeah it's a it's a international standard and framework for security and it's O wasp yep o wasp top 10. yeah yeah I mean that would definitely I can imagine how that would be a huge concern recently from government you know like critical infrastructure stuff that's changing around and obviously the Optics and things that have happened recently so no doubt you get that question a lot so it's good that you have that as a a front of Mind topic the other thing I want to touch on was Talent like it's we've got um pretty big culture here at Red to try and keep our Engineers happy and that kind of thing software developers are probably similar but um or maybe even harder to find especially with like a lot of experience how do you attract and retain Talent yeah totally so um there is a number of different ways so the first way is that we're heavily involved within the local University communities um so we sponsor a number of different University groups so uqcs uq Computing society and we're frequently going there and we're working alongside those what we also do is we invest in our people so in the devops center for support we'll actually put undergrads on um before they've graduated and we hire probably 10 to 15 per year to go into the devops center on a casual basis they can come in they can learn the ropes learn the technology everything like that as well and then when they're ready and they've graduated we can promote them to be a developer at that point once they've been exposed and have that experience so some of the best developers and Squad leads that we've had have been within the business for five to six years and they came through that pathway right from the University which is fantastic in terms of how we attract Talent that's already in market so we have an internal recruiter we reach out to people that we feel would be a good fit and then we actually put them through a four-week training program so we call it bot Camp to learn the technology and learn our process before they're Ever Getting nearer project or a client or anything like that as well Judy pretty vigorous testing on people before you harm do you get them to actually go and build something or show them stuff you've built before hiring yeah yeah that's it so there's a few exams um they go through one or two exams and they could build something and they can also bring a long um a project that they've done and things like that as well so there's it's very rigorous yeah yeah so it sounds like um uh with the cobots and that's like consistently developing but what other technology are you developing a working Mouse uh it all depends on the client's need so in terms of projects that we're doing we can do iot projects and hardware-based projects as well but it all depends on what's the best solution for the problem so it's a hard one to say the thing is we use the right tool for the situation so for example we have previously developed a music platform for commercial music streaming which I know your library and I do yeah um so tell me more yeah so uh think it think of it as Spotify for a commercial music venue so they've got the right license to play the music but not only that but it also is tailored to the the venues brand so the music's playing that matches the venue so it's it's improving upon the experience so they were previously licensing a piece of software from the UK very very old paying a lot for it and they just signed some big hotel clients and they went right well why are we licensing this let's build it ourselves and let's use that to deliver so in that case there was a big web component so there's a web DJ component for basically creating the playlist setting up the venues doing the music and the moods and everything like that but we need to get players with the music actually on premise so the way we did that is that we had the web app for the player we then have a a local application installed you on the actual player on site with another application on that and that application was using a technology that could make it a desktop application so it just appeared like a native desktop application and then there was a mobile app so there's a mobile app to actually allow managers in the venue to get like the music just like the music Control the playlist although there's a kid's birthday party and explicit songs just come on I'll skip that track and they have that kind of control so what can we know what that is oh I can't say that one's a top secret yeah so it is it it is cool um good jacket like you must work on some cool some cool projects and you must have quite a range of clients so is there a typical client client size you know if somebody wanted to engage working Mouse is there kind of you know the project needs to be of a certain size do you do small projects big projects yeah what's the I don't know if there's a typical but yeah maybe talk about the kind of projects yeah so the different groupings and categories of projects so in terms of domain um so we are completely domain agnostic um so we can be working in finance we can be working in music we can be working with mining um we don't mind um we are the experts in software development and we will learn the domain and the product owner and the customer usually brings that domain experience and for us to be able to solve the problem we have the software experience in regards to type of projects um does a grouping of projects Greenfield and Brownfield so we can adopt both Greenfield and Brownfield when we take on a Greenfield project that means you're starting from fresh it's a new problem with a new business process and we haven't done anything before great Brownfield is I've built my application with a single developer there's no documentation please help me um and understandable there's lots of people that are in that situation and we can take those those on as well but it's a bit harder and we leverage a bit of the technology to relearn and rebuild that there's different modernization options that we have for those projects and it's Case by case if that makes sense and you mentioned iot before like do you get involved in Hardware designing Hardware yeah we don't do your Hardware design we're definitely not Hardware experts and we're not um we're not experts on software on Hardware devices as well so it's mainly Integrations so firmware we don't do firmware yeah so web software mobile apps that's our core cool so someone would come to you with a device they'd have the firmware or that those bits and pieces and you would develop the front end effectively for that yeah exactly so like um another example of one is there is a um steel manufacturer in Australia and they have a paint production line and they comes out on a big reel and if there is a defect on that line each wheel is worth about 200 000 so how do they know if there's a defect if they're only testing the end of the line and it's right in the middle of the line so we've built them a product that enables the image to be captured of every meter of that line and then they can go back and review it so when a customer says oh there's a dent in this we can go back and go no there's actually there was no dent in that or they can observe a fault on the line immediately and that's linked up on a local on-prem piece of hardware and there's cameras so they are the experts in that and linking up with the plcs they purchase the cameras and they're in control of the cameras we created the software to talk to the cameras so we we took the camera's firmware and we were integrated with the camera's SDK in its firmware so are you doing Ai and image recognition and those those kinds of projects yes we do a fair few kind of machine learning and AI projects we've done some in finance and we've also done some in Machine Vision as well and it's amazing what you can do I totally agree with why Tesla is going down the Machine Vision Route yeah we've we've looked at some things in the past and um it is amazing what you can do so it's such a cool space um we must you must see you must see some crazy ideas come across the table as well some crazy business cases that you're like I really want to prove this you have to um you have to bring people down to reality sometimes if someone comes with a whack it you're like that's just no that's not gonna work yeah there's a few a week oh really it's like um it's difficult I mean you can do so much with software but it's process automation there needs to be a process behind it for you to build um and there's always lots of ideas of the next Uber of or anything like that yeah we used to joke um within the company that we'd never tell Uber drivers that we were a software development agency um and because you go I have got an idea for a nap luckily these days you can because they just won't talk about that crypto which is fantastic yeah yeah but one thing I want to bring up um I think there's a maybe it's just me I think the increasing demand of this the question right now in the market right now so say say there's someone out there who needs something developed and they're looking at do I hire a developer internally as a business or do I Outsource to an organization like working Mouse I think I come across that a lot um I don't know if you have what are some some takeaways and things that you would say just consider this if you're looking at getting someone internally versus Outsourcing yeah definitely um it's a difficult one um it all depends on again the business case and where they want to go with it um so uh if you if the consideration to take into place if you're going to create your own development team is that a single software developer is probably not the best way to go about it because they'll just build it you won't have your documentation you won't have your testing you won't have your quality so you need at least two software developers to be able to peer review and then you need someone who understands the process and understands what the quality needs to be um to be able to actually make sure that they're doing the right thing so you need someone to manage them you need a squad lead um to follow an agile style and then you also need someone who's the product owner to direct the team so before you know it you've kind of built out a few different roles there and then you need the process for them to follow you need to agree on a tech stack and you also need to make sure that there's a certain level of experience within that team so one of the biggest things I love and works so well within our organization is that because we've got 70 people someone's done it in some form somewhere and they can share that knowledge and they're not just working within a silo that's really good now can do they want to do we want them to carry on with us forever we're completely open to them creating their own team so we've had about four customers who have previously created a product with us and then we've helped them train up a team and they've taken that team and they're continuing to develop that as a product as well I think that's a good point like there's multiple skill sets I mean for what we do for what you do there's multiple skill sets that are required to get you to that end result it's not just one size fits all there's not one person and I think sometimes that's probably overlooked it's just oh it's just software development yeah a developer and they'll build that thing for us there's multiple roles and multiple skill sets multiple different personality types along that journey to be able to get you a working solution yeah without backing themselves into a corner and also we've we've made a lot of learnings so we say whenever we've made a learning on a project we go where do we stuff up what was the root cause of that we have an analysis and we do a retrospective on it and then we go right what do we need to change with our process so we we mitigate that in the future so we've got that sophisticated way of working where we've been doing it for 10 years and we know right this is how we identify risk and we manage risk in a software development project that's really cool so what's next for working Mouse and I guess after what you've done now you've got a lot of things on the go but as an organization what's next um so next for us um is um at the moment we've just finished a big r d round um with the technology set and kobots um and we're taking that technology set and the fourth generation Bots we're calling them to Market um so we got a we got a lot of good feedback from the market when we launched the third generation Bots with kobots and everyone said this is fantastic I love what it can do but can I use this technology set instead of this technology set because we decided a couple of years ago as an organization that that's what we're going to do we went no we we actually control the Bots and and they went okay it's a deal breaker for us so we've just figured out how to actually train a bot and release that as a product so what that means is that you can train a bot on any technology stack and also any problem set as well so it doesn't need to be just for software development we can actually train a bot with a custom meta model to effectively create automation wherever there is a pattern in anything so taking these fourth gen Bots out to Market I'm intrigued and I'm confused yes that's half the problem is explaining it and people's eyes not gazing over what's an example of a real life example what's an example and who is the customer so is this an and you are you selling to software developers are you selling to businesses are you selling to so yeah software developers so um they can create their own bot so different agencies but also people that understand and do software development on a daily basis so they don't necessarily have to be software developers but they have to understand it right so improve their workflows improve their efficiency it's all time-based effectively so if you can do that task in half a time whatever it is exactly right cool so there's a there's a concept um that we call judoka and I'm probably massively slaughtering that works it's Japanese Toyota production system um and what that means is that you you build something and then you automate it and then you have control of that automation for when it fails so that concept we're bringing to Market with devops and these fourth generation bots so a use case is that there is a government customer who has a legacy system that Legacy system has over 300 Integrations and each integration would take a development team approximately one iteration so one to two weeks to do so that's a lot of Labor um to do that so we built a bot specifically for managing the integration from that Legacy system to their their data integration layer um so that can get basically each integration can get built in a couple of hours and deployed so instead of someone actually like developer going in building that integration 300 times they get you pay for your Bot as a service and go and plug that in make it work and it just learns the way that it doesn't matter what they've built it on and it goes and learns how to integrate that yeah we have to trade it so we train it um on the best best practice and there is a learning algorithm in there that says is this the right thing is this a variable here etc etc or the static and we go no this is right this is right and then we deploy that to the platform and that bot now has learned that behavior if that makes sense and then we can reuse that again and again and again makes sense yeah all right thanks Dave really appreciate you coming you've given some really good insights into if you're looking at um right I've got another question yeah so you've been doing this for so many years what's your favorite project okay a tough one um you can you can do more than one if there's a tie yeah that's it because they do different things right yeah um so the music one I was saying before I absolutely love it's pretty cool we'll talk about that after that yeah because it kind of it hits so many different technology stacks and delivers a really good service um but there was a product we built a better part of seven years ago and I still love that product and the customer didn't end up using it um and I I thought it was better than MailChimp okay and I was like this is awesome it was a forms based product and then enabled them to basically build their own custom surveys questions and send it out um I was like this is better than MailChimp I didn't use it oh it's just so what happens in that case so they just just dies a slow death or yeah that's it um that's why you need a strong business case yeah um and um yeah the customer owns the IP yeah so it's it's your IP you do with it as you will um if you want to take it off and start developing it over here like that that's absolutely so what they did did they just buy MailChimp or something else I don't know yeah thanks I really appreciate you've given some really good takeaways from organizations about we're looking for something to be developed that um what to look out for some horror stories and some of the Innovation you're doing working Mouse with kobots is is really cool if anyone needs to reach you how can they how can they reach out yeah just uh jump on the working Mouse website www.workingmouse.com [Music] foreign [Music]

2022-10-27 12:11

Other news