Podcast 009 – Jordan Kyriakidis – Founder & CEO of QRA Corp

Jordan Kyriakidis – Founder & CEO of QRA Corp

“QRA Corp’s mission is to accelerate the design process across industries who are tackling the most complex systems by empowering them to build tomorrow’s safe, secure, and incredibly powerful products. QRA’s technology, patented toolsets, and capabilities have been used to avoid stressful reworks, enable confident engineering, and find previously undetected catastrophic flaws.”

Show notes

Discussion with Jordan about the current state of Systems Engineering and where it can go in the future. Before founding QRA Corp, Jordan was a professor of quantum physics at Dalhousie University.

Jordan’s web links:

Time stamps:

0:00:00Introduction
0:00:56Jordan’s Background
0:01:24QRA is centered on requirements and improving requirements
0:07:56Entrepreneurship background
0:12:01Current state of systems engineering
0:15:44Impact of quantum computers
0:17:49Rise in AI
0:22:06Industries in demand for systems engineering
0:24:23Culture differences in various domains
0:26:29Processing requirements in different languages?
0:28:01System Engineering in the next 10 years
0:33:37Largest semiconductor manufacturer Nvidia
0:35:25Demand of quality products from the consumers
0:36:40Advice for novices in system engineering
0:39:40Design level thinking
0:41:31Jordan transition from being a Professor to CEO
0:45:51Career opportunities at QRA Corp
0:47:53Impact of Covid in day to day business operations
0:53:22Conclusion

Transcript:

00:01[Music]00:15um today’s guest00:16is jordan kiriakidas who is the founder00:20and ceo00:20of qra corp a company based in nova00:24scotia’s canada halifax00:28and we’re going to talk about uh what00:30their company does what it offers to the00:33systems engineering industry and also to00:35their individual customers00:37and learn quite a lot about um jordan’s00:40background that led to him to to create00:42this company so00:43welcome jordan thank you joshua’s uh00:46pleasure to be here00:47jordan could you give us a bit of an00:49overview of uh your company qra00:51and what does the acronym stand for sure00:55um the acronym doesn’t stand for for00:58anything um it’s just the it’s just the01:01initials the q stands for quantum01:02as a um as a nod to my to my background01:06which01:06i assume we’ll get into um what we01:10actually do right now we are01:14[Music]01:15we are a technology software company01:18um and we sell primarily to enterprise01:21and engineering organizations01:24and we currently we solve three problems01:26for our customers and they’re all01:27centered about requirements and and01:30improving requirements01:32um one problem that we help solve is a01:34lot of our customers01:36they undergo difficult requirement01:37reviews a very costly01:40very timely and the outcomes aren’t as01:42well as you know01:43they want them to be a second problem is01:46that really01:47some of our customers lack an01:48understanding of what are requirements01:51best practices right01:52they know that requirements need to be01:54improved but they don’t have a good01:56understanding of what are the best01:57practices and how to verify that you are01:59adhering to those02:00best practices and the third problem is02:03just a general02:04improvement of requirements quality02:08especially if requirements are going to02:09go on you know the typically they’re02:12they are the initial phase of a product02:14development or systems development02:16and they know that poor requirements02:20quality02:21will have compounding dangers further on02:24in the life cycle of the02:25development process and so they really02:27want to improve quality02:29you know almost all requirements today02:31are written in natural language02:33in fact most of the world’s information02:35is written in text so natural language02:37processing02:38and natural language understanding are02:40the appropriate technology to use here02:43so what we actually do with our software02:46tools is02:46provide a quality score for each02:48requirement that you have02:50and also for the collection of02:51requirements uh as well02:54we would look for uh term consistency so02:56if you call something02:58you know a rover a module a vehicle03:02different the same thing different terms03:04in your document we’ll flag those and03:06say maybe you should03:07you know consolidate the terms check for03:09units consistency03:11we can look for uh requirements that are03:13very similar03:14which often means they contradict each03:15other um so those types of uh03:18um solutions we provide to our customers03:21okay03:22so your what would when when qr03:26a company is turning up to a company or03:29a project03:30what would it look like um from day one03:32are you consuming03:34requirements that already exist in in03:36word documents or03:38other requirement management tools yeah03:41so03:42uh it is it is a product that they would03:43buy and install on their premises and03:46if they have requirements written we03:48have several integrations word in excel03:50um so many even for large projects start03:54off03:55in either word or excel and so we03:57provide an integration to that so that03:59you don’t have to move your requirements04:00anywhere else you can analyze them04:02wherever you are04:04also many of our customers use04:07specialized requirement management tools04:09like doors which is04:13a product by ibm there’s polarian which04:16is a product by siemens04:17there is a jama connect04:20by jama right so if you use those our04:24qv scribe is the name of our product it04:26integrates directly with these tools and04:27you don’t have to move your requirements04:29anywhere you can just stay04:30in place we provide this added04:32functionality04:34on top it’s kind of a semantic layer on04:36top that’ll just analyze04:38analyze your requirements so you can04:40analyze requirements that exist but also04:42if you’re authoring requirements04:43we are our products right there beside04:45you uh advising as04:47as you type the requirements on on the04:49quality of them04:51so is it kind of almost like you’re04:53adding like this layer of04:55a skilled requirements engineer who04:58is analyzing either requirements that05:00you’re generating or05:02existing ones and then within whatever05:05existing database you already have05:07and as you mentioned improving their05:10quality or05:11guiding engineers to be able to improve05:13their quality that’s right05:15in fact a lot of our customers have you05:18know the reason05:19they want to improve requirements05:20quality for example is that05:22they may have a lot of engineers writing05:24requirements that are very young05:26and so don’t have the experience or05:28english is not their first language05:31and so typically they would be paired05:34with a senior engineer05:35and there is a lot of back and forth uh05:38on improving the quality of the05:39requirements you know they may not be05:41atomic05:42uh they may have excessive continuances05:44and so05:45you know the senior engineer would05:46advise them and tell them to break it up05:48and the issue they have if they don’t05:50use our tool is often times they’ll go05:52into a review meeting05:54and they end up talking about the syntax05:56of the requirements05:57as opposed to whether the requirement is05:59is correct right so06:00using our tool allows them to have the06:03discussion at a higher level06:04right what they’re talking about is this06:06the right thing that we want to build06:07does this actually capture our intent06:09instead of no what do you mean by this06:11requirement or this seems to contradict06:12that requirement so it’s a higher level06:14of06:15discussion that can they can have and06:17are you finding that the06:19demand for this type of product or06:22service06:22is is on the rise that people talk about06:24how systems are getting06:26ever more complex and it sounds like06:29that would only mean that06:30you would need these types of tools even06:32even more06:34uh absolutely yeah the demand is06:36increasing uh06:38a tremendous amount and and we can get06:40into why that’s increasing but that is06:42absolutely true06:43uh in fact the first product we built06:46was06:46design tool to make sure that the06:48designs were correct right and what06:50ended up happening06:51is most of our customers06:54ended up comparing their design06:56implementation with the requirements06:58and uh what we07:02found out is most the requirements are07:04very poorly written right and everyone07:05agreed that they’re poorly written07:07and so uh you know i’m a physicist by07:10background so i like by nature like to07:12get down to first principles07:13in this case it’s finding the error as07:15early as possible and07:17we realized fairly early on that07:19actually07:20you can make mistakes even before you07:21get to the design process when you’re07:23first07:24capturing your intent and writing down07:25the requirements so we should focus07:27we should focus there and uh and that’s07:30where we saw a great success07:31both in commercially as a company but07:35um also the value we provide and the07:38solution provide our customers07:40okay yeah so how did the company07:43come to be created i knew you were a07:44professor at um07:46dalhousie university in uh quantum07:48physics07:49and so as you’ve you know mentioned into07:51requirements or tools to07:53to improve requirements seems like a big07:55leap07:56it is so i am a bit of an accidental07:59entrepreneur08:00uh the the plan wasn’t to do what we’re08:03doing it’s just following my nose and08:05following the interesting uh problems08:07right like as you said08:08the uh i am an academic by by training08:11and an academic running a company i08:14would say there’s pros and cons08:15associated with that but my background08:18is in theoretical physics08:19uh quantum theory in particular quantum08:21computing quantum dissipation08:23um that was kind of my bread and butter08:25and you know i was08:27for many years um i was on 14 years or08:30so i was on the path to being08:32the venerable theorist with the leather08:34elbow patches and the whole bit08:36but then i ended up getting a r d08:39contract08:40on quantum computing with an aerospace08:43company08:44and that contract was08:47was very successful and08:51you know the value of the contract ended08:52up increasing the team08:54expanded to service the contract08:57but it became progressively more applied08:59so we went from doing the fundamental09:00science09:01and then you know from there i09:03progressed to understanding the09:04engineering problem they were trying to09:06solve09:06and so it became a bit more applied and09:08then from there i went and i learned09:10the business problem that they were09:12trying to solve and it became even more09:13applied09:14and as we learned more about but the09:16problem the way we solved it was09:19was um you know kind of veered the09:22technology we use the types and the way09:24we solve the problem09:25um changed as we went along and we ended09:28up09:28at a point where i thought okay i’ve09:32kind of deviated quite a bit from09:33physics09:34the problem is very interesting and09:36these systems these engineer systems are09:37becoming so complex09:39a lot of the principles that we have in09:41theoretical physically developed over09:42the last09:44half century one century can be applied09:46now to these synthetic systems09:48because they’re so complicated so many09:49interacting pieces and so we started09:51making progress with that09:53it got to a point where i had to stop09:55developing algorithms and start09:57developing a product09:58in order to get the engineers to10:00actually use uh the algorithm to10:01validate them10:02and then we had to get a company to10:04distribute the product to get it to the10:05people10:06that’s kind of how it started so the10:07first you know it kind of started10:09backwards i would say10:10uh than what is typical um and then10:13after we did that10:14then you know we started learning about10:15how to run a company what what that10:17actually means and now you know that was10:18that was quite a year ago that was kind10:19of that was quite a while ago i should10:21say10:21that’s kind of our genesis of where we10:23are of where we are today10:25and when was the company founded uh10:28it was company was founded in um10:32late 2012 around there so we’ve been10:34around for a while10:35we had various different iterations we10:36had you know we had multiple products10:39and went through several phases we10:41started off as10:42as doing mainly large contracts um10:46and and a lot of services um but now10:50we are primarily a product company we10:52just sell we just sell our product and10:54um10:54and you know that that’s how we operate10:56now that’s great yeah be great when you10:58hit that 10 year 10-year10:59anniversary uh next year yeah we are uh11:04we are uh the social committee is11:05planning the party already11:09so you mentioned that you’d sort of work11:11backwards11:12and you were almost developing a product11:14to solve your own problem when you were11:17running off physics project i mean how11:20do you see11:21uh which perhaps gives you an image of11:24a lot of the systems engineering life11:26cycle but how do you see the current11:28state of11:29um systems engineering in general11:31because i imagine through all of your11:33your uh customer interactions you have11:35quite a good um11:37view view of yeah and the fact that i am11:41not a systems engineering by training so11:43i had a view from the outs but i am a11:45technical background so you know i can11:46understand it11:47um that sort of gives me a different11:49viewpoint on it11:50and so to my mind the current state of11:53systems engineering11:54is that the fundamentals are are quite11:57strong11:57right the principles in which they’re11:59founded are exactly the right principles12:02the uh the12:05systems engineering as a distinct12:07discipline is gaining in importance all12:09the time and will continue to do so for12:11decades into the future12:12there’s no question of that at the same12:15time there is a great deal of change12:18and evolution even disruption disruption12:20coming12:22one of them is is um12:25i think it’s kind of three uh items in12:28particular one is technology12:30so new tools are are available now that12:32can automate12:33uh parts of the work especially design12:36right we had a big12:37cad for example is a huge change in the12:39way12:40the discipline works now the tools and12:43technology that’s present today12:44is i think we’re near another kind of12:46inflection point12:48another one is market demands products12:50themselves are becoming far12:52more far more complex and so too12:55are the methods and techniques of design12:58and manufacturing right so even though13:00your product may be simple the way you13:01manufacture is becoming13:02increasingly more complex and software13:05is eating the world right13:06software is becoming far more important13:10and the third one is is people right13:13the um experienced engineers13:16are retiring and young ones are moving13:18in so there’s kind of a generational13:19shift happening now13:20happening now as well so because of all13:22these things i would say there’s a13:23mix of anxiety and confidence anxiety13:27because there is a change coming it’s13:28not quite clear what the other side will13:30look like13:31but there’s also confidence because13:33systems engineering is gaining an13:35obvious and recognized importance13:37right appropriately so yeah and i feel13:41almost like this current pandemic has13:44to a lot of governments around the world13:47the idea of sort of thinking about their13:50country as a system or the13:52the globe as the system and how um not13:55only sort of engineering of it but how13:57things do13:58interact with each other and how14:01ideas of resilience and these sort of uh14:04various quality attributes that we might14:06have in our papers that people14:08say they care about but then don’t14:09actively act on14:11um i feel like that might have been sort14:14of14:14hopefully expedited some of the ideas14:16that in systems engineering we14:19have written a lot about in books but14:21maybe haven’t been able to sell to the14:22wider public14:24i think there’s no question of that14:25right and in fact that’s something that14:27resonates with me a lot14:28because my you know i was my background14:31in physics in particular14:32is theoretical condensed matter physics14:36and that’s all about interactions of14:37strongly interacting systems14:39and they’re the whole name of the game14:41like in systems engineering is14:42it’s really everything is in the14:44interactions less about the components14:46themselves right so14:48we have this concept in physics and they14:49do in engineering each component14:51could be exactly the spec could be14:53working properly14:54but depending on how they interact the14:57whole system can have new14:58emergent properties that do not live in15:00an individual component15:02it’s the interaction that interaction15:04that matters so you need to15:05elevate the interactions to first-class15:07citizens15:08and that gives you a new viewpoint on15:10how the system operates15:13yeah it is very fascinating and i kind15:16of wanted to switch gears into your your15:18background as a quantum physicist15:20so i’m not sure how many people15:24on the podcast sort of know about the15:26potential of like quantum computers and15:28how that15:30might sort of impact um how we do15:33system centering design or or other15:35aspects of system surgery15:37as an expert in quantum like where are15:39you seeing15:40that impacting how we can do systems15:43engineering15:44i i think quantum computing in15:46particular um15:49i would say the jury is still out and15:51what i mean by that15:52not whether it’s going to affect it or15:54not i think it clearly15:56is the time frame of when it actually15:58will affect it is not15:59clear but the technology16:03clearly has the potential to be a game16:05changer what i mean that’s unclear is16:08how in particular is going to be used16:10right so16:11one of the analogies i like to16:15um use here is if you think about the16:18laser16:18the laser is just light right so but16:20when the laser was16:21first invented or think of the16:22applications16:24you know we don’t use lasers for light16:27bulbs even though it’s just light16:28we don’t use it for illumination we use16:29it for all these so many other different16:31applications16:32that come from lasers even though it’s16:33just light i think quantum computing is16:36going to be16:36a lot like that right it may be used in16:39the design16:40but it’s going to be used a lot for16:41example in sensors for sure16:44in some processing and some ai systems16:47especially training of models could be16:49used quantum16:50quantum computing cryptography obviously16:52and security16:53is obviously going to be going to be a16:55big deal so it’s going to be somewhere16:57in there where these systems are going16:59to play so we should make17:01as you mentioned like a laser we treat17:02it as we’ve got this new component17:04and once that component is um in the17:07state where it can be commercialized we17:09might17:09a whole lot of applications that we17:11don’t even know about yet17:12that’s right that’s right yeah i don’t17:14know like17:16cds i don’t know who at what point from17:19laser invention they were conceived17:22that’s right if it was prior17:23yeah that’s right so i know my own view17:25and this is speculation17:27my own view is that it’s not a quantum17:29computer is not going to be17:30a computer like we have now that you17:32type on it’s just going to be a special17:34kind of processor right17:35i mean that might happen to me i think17:36that’s unlikely it’ll be maybe more17:39specialized processors17:40uh that live and that live in different17:42places17:44so i guess with your company today you17:47mentioned natural language processing17:49and17:49ai would those be the most interesting17:51new technologies that you would be17:54actively engaging for for your products17:58yeah today we are yeah a lot of ai18:01and a lot of natural language processing18:04and natural language understanding18:06only because that is18:11where the requirements live right it18:13lives in text if you want to analyze18:14them18:15um you can you need to use natural18:18language understanding right18:19you can then go beyond that you can18:21think of what happens afterwards like18:23from you know can we generate18:24architecture diagrams uh systeml18:26diagrams from the requirements directly18:28those systems can help that18:30um or if you have certain requirements18:32and have a model18:33you know you can use ai to tell you what18:36parts of the model or suspect that you18:37may want to look at right so there’s18:39those elements of of ai as well18:42we don’t do much of that right now uh18:45right now is primarily natural language18:47processing because that’s that’s where18:48we are and the technology in that is18:50increasing so fast i think microsoft18:53using gbt3 has just developed some tools18:55where18:56they can change they can go from text to18:59code right19:00it’s been a dream for a long time to be19:01able to do that i think we’re19:03we’re uh getting much closer now than19:06we’re than we were before19:07okay so i guess you can piggyback on the19:10advances from19:12uh what’s going on with companies like19:14microsoft and google who19:16seem to really pushing the envelope in19:19those algorithms19:22yes to a certain extent to a large19:24extent we can um we don’t need to19:26reinvent the wheels19:27the difference is if you look at19:30engineering requirements or19:32technical specifications more generally19:34let’s say19:35they tend to have a very particular kind19:37of language where most of the general19:39models19:40uh produced by the big companies are19:43trained on like newspapers wikipedia19:45they’re just19:46general pros right and and so it’s a19:49good base but we need to do some19:50training on top of that to you know19:52steer it towards19:54the technical language and if i find at19:58least19:58a lot of companies and organizations20:00they’ve almost got their own language20:02that they have sort of20:06it’s never been sort of codified or20:07taught but it’s it’s become that certain20:09words mean certain things20:11in certain organizations and they mean20:12different things in different20:14organizations20:15they do yeah that’s right so some of20:17them have you know like there’s a20:18difference between the word20:20must and the word sha right they they20:23mean different things and they codified20:24it so that’s that’s one of the example20:26that’s one a good example that we can20:27help with so they’re codified20:31most companies today verify that20:35manually right meaning they have an20:37engine typically a senior engineer is20:38one of the most expensive resources20:40their job is to read the requirements20:43with a highlighter in hand20:44and you know circle all the errors they20:47see and until20:48their eyes bleed and so a lot of the20:50stuff can be automated20:51right so with your um20:55sort of research and development are you20:57in are you working with universities or20:59can you do all this in-house or21:02we do we do most of it we do most of it21:04in-house21:06well we do most of it the product21:09development is all done in-house21:11we work a lot with our customers we do21:13work21:14less so now with universities than we21:15did in the past well we do have some21:17university projects21:18these are mainly far-reaching21:20far-reaching things21:22um really where we get the most insight21:25to be honest21:25is working with our customers because21:28then because we’re a small shop21:30and we have competitors and our21:32competitors are our big companies21:34um we can offer um21:38a much more insightful and specific um21:41solutions to them and for that to work21:44we need to really understand our21:46customers very well and so we’re21:48oftentimes talking with our customers21:51and developing with our customers21:53to figure out what actual problems they21:55have and that’s probably the source of21:57where most insight comes from21:58from our customers mm-hmm that sounds22:01very22:02sensible and seems like it’s worked very22:04well given you’re coming up 10 years22:06so far so good yeah so you’re working22:09with customers offer22:10a wide range of industries22:14yes um primarily22:17it’s uh oil and energy companies medical22:20device companies are probably our two22:22biggest industries22:23uh aerospace and defense companies are22:25up there22:26and automotive uh and semiconductors are22:29kind of sort of22:30our industries right as you may not be22:33so surprising those are industries that22:35really value systems engineering they22:37have products that need to adhere to22:39some regulatory standards22:42and there is a criticality associated22:44with them22:45like you know it’s important for them to22:46be correct and22:48it’s expensive to develop and expensive22:50to change once they find errors22:52down the line kind of saying22:56i’m just going to say as opposed to22:58saying a typical software startup that23:00if they find an error they can just fix23:01it and you know no one dies if it goes23:03wrong right23:04yes it’s like a quite a different23:06culture um23:08from i don’t want to say silicon valley23:10because i think that’s that’s a bit23:11unfair because in silicon valley you do23:12have the intel’s and23:14um companies who do care a lot about um23:18[Music]23:19zero failures um but but it is23:23maybe a little troubling that there was23:27perhaps since the year 2000 or so when23:29the big internet bubble where23:31there was a more focus on um speed of23:34execution as opposed to23:36avoiding failure and that’s right yeah23:40that that’s that’s very true and it23:43depends a lot on the company right23:44so you can think of23:48very crudely speaking two groups of um23:51technology companies one that came from23:54pure software23:55and are evolving into hardware um sort23:58of one23:58culture let’s say it’s a bit unfair like24:00you said to categorize it that way24:02the other one is from um24:06um mechanical and electromechanical24:08systems24:09and like electrical engineering and24:10mechanical engineering that are now24:12becoming more software24:13eventually everyone’s becoming a24:14software company but depending on where24:16they came from they have different24:17levels of24:19rigor and tolerance to error in their in24:21their work24:23yes and do you find but within the um24:26domains you do work24:27that there are differences in this the24:29cultures or24:30the types of solutions that they prefer24:35um not so much actually so it’s not uh24:39it’s not that we have to specialize to24:41one industry24:42or the other um the in the end24:45they are all systems engineers and there24:49is a corset of principles that have to24:50adhere to to make these complicated24:52systems that work24:53is a combination of software and24:55hardware and there are a lot of24:57interacting components and there’s a24:59supply chain and there’s25:00different people working on it so25:02because of those things are the same25:04the technical issues are the same25:08and the solutions are the same where25:10there are differences25:11in is more on the business side25:14so a company in the medical devices for25:18example25:19their main drivers may be time to market25:21so they want to make sure that the25:22requirements are correct25:24because it’s a race to get something to25:25market because their competitors are on25:27their heels and generally they’ll have25:28a product they release and then they’re25:30working on the next one whereas someone25:32say in aerospace25:36their main driver is correctness things25:38have to be correct and they’ll take the25:40time to make sure that it’s correct so25:41the requirements25:42have to be correct because they cannot25:44tolerate any errors right so25:45it’s more than business drivers and the25:48structure of the company whether it’s25:50a set of independent teams or whether25:53it’s a more top-down structure25:55we find those differences um among their25:58customers25:59not so much the core technical issues26:02okay26:02that’s interesting and do you um to work26:05with26:05uh clients um outside of north america26:10we do we have some in yeah we do we have26:12some in germany we have some in uh26:14india i believe some in the uk as well26:18most are in the states and a lot of them26:20are26:21though multinational companies26:24oftentimes are on call where people from26:25different26:26you know different countries are on the26:28call26:29and are you processing requirements in26:31different languages outside of26:33english no now it’s primarily english we26:37we26:38didn’t get asked recently to if we can26:41offer a translation to korean26:43um we are not equipped to do that now26:46luckily most of the requirements are26:47written26:47in english um and the26:51issue with doing things in different26:53languages26:54is a lot of the language models a lot of26:56the nlp and ai that we use26:58are trained on english language right27:01and so27:01it’s not a matter of just localizing the27:04software like27:05the menu items and stuff like that it’s27:06the core algorithm that are language27:08dependent right and so27:10right now we have more than enough work27:12in english but27:13[Music]27:15doing things in a different language is27:16in our future and i imagine27:19jumping straight to korean is maybe they27:21talk about languages similarity and27:24i don’t know maybe english to french or27:26english or german or something27:28yeah i would say probably the next yeah27:31that’s true27:32the next language to do germ germany or27:35german27:35probably be the most the most likely27:37maybe french27:38um it depends on big automotive uh27:41um how how big a sector automotive is27:44gonna be for us27:45yeah yeah so i guess if you’re coming up27:49on your 10-year27:50anniversary for for the company where do27:53you see the the next 10 years for27:56your company but also um systems27:58engineering27:59as an industry um28:02if i look at systems um28:08yeah so i think it’s um28:12again this is largely speculation28:15and if we look at for example28:18so there’s two ways i cannot answer so28:20one is you know what is the uh28:22the state of requirements writing how is28:24that going to change over the over the28:25next or the next decade28:26and then there’s a broader issue of you28:29know28:30what systems engineering going to look28:31like uh in general28:33um so maybe if i talk about requirements28:36first then we can veer off to our28:37systems engineering a bit more broader28:41you know we talked earlier about the28:42state of systems engineering today and a28:44lot of over the next uh28:46the challenges facing requirements28:48engineering today in the near future28:50really are outcomes of those you know28:53one28:53thing that we see different now is that28:56most companies28:58have either moved or are moving or plan29:01to move from the traditional waterfall29:03manufacturing technique into a more29:05agile and flexible approach and so29:08requirements are still vitally important29:11because you need to capture your intent29:13of what you want to build29:14but whereas maybe in the past29:17requirements were29:18more uh in principle one and done like29:21it’s a very distinct phase when you’re29:22done you move on to implementation29:24what we’re seeing now and continue to29:26see is requirements like the design29:29will um like the design and the29:31interfaces the functional and physical29:33components all this information will29:36in parallel mature over time right29:38there’ll be several several passes29:40and so that means that it’s more29:42important that as this information29:44matures29:45you can maintain coherency and coupling29:47between all these different artifacts29:48the design the requirements the29:50interfaces29:51today this is done largely manual and29:53the burden is becoming just too great29:55for you know individual individuals to29:57do that it’s not really a30:00a question of the engine the caliber of30:01engineering it’s more a question of30:03complexity30:04and one putting all that information30:06into one head30:07i always say that’s one big thing that30:10both tools and processes are moving30:11towards30:12it’s gonna take about a decade before30:13that process is complete30:15and the other one is technology30:20itself right and requirements are30:22becoming more important now because30:24you know for a long time this is still30:27true today but for a long time30:28most new products were incremental30:30improvements over the previous versions30:32uh and so were if you want to improve30:34the process they were incremental uh30:36improvements so the engineers that were30:38building the products30:40um they were uh you know they were they30:43were pretty smart30:44if they saw the requirements were bad30:46and they were bad30:47you know the engineer could say you know30:49i don’t know what kid wrote these but i30:51know what i’m doing because i’ve built30:52these products before and i know how to30:53build it correctly30:54and they were largely correct what’s30:57happening30:58today because there’s new technology31:00sometimes disruptive technology31:02[Music]31:03they’re being introduced deeper into the31:05products31:06mainly sophisticated software like31:09sensors into the controllers into the31:10actuators they’re all becoming software31:12enabled31:13and so uh engineers have less experience31:15with these new non-incremental kind of31:17uh31:18systems and so because they don’t have31:21the experience they rely more on the31:23specs and on the requirements31:25so even though the requirements you know31:27i would say were never really all that31:28great31:29um their importance is becoming uh31:32greater as time goes on because people31:34are relying on them more and that trend31:36is gonna is gonna continue in the future31:38and so things like the the quality the31:41completeness and the correctness of31:43requirements are beca are gaining a new31:45importance now31:46then because they’re relied upon now31:48more than more than ever31:50yes and i think you’ve touched on the31:52retirement of engineers so31:55even if we were we weren’t radically31:57changing31:58the various technologies in it32:01your people will retire and if you’ve32:03lost that experienced person who was32:05catching a lot of your32:07your problems um how long can you you32:10last you can’t32:11keeps on working forever that’s right um32:14that’s right so there are multiple32:16issues coming in here for the32:18human resources and the personnel32:19problem as well as a technological32:21problem right32:22and i think that’s also true for systems32:24engineering in general32:26but i think over the next say 10 years32:28or so there’s gonna be lots of changes32:30on systems engineering32:31itself um i think there are32:35the way you mentioned quantum computing32:37that’s going to have an impact but i32:38kind of see32:40three main technologies that really are32:42just starting now32:43they’re going to continue to spread one32:45is the whole idea of adaptive algorithms32:48that we’re going to have algorithms in32:50our systems that the algorithm itself32:52changes as as the more you operate the32:54piece of equipment or the vehicle32:56another one is uh generative design33:00and the third one is additive33:02manufacturing so these three33:04manufacturing technologies are gonna33:06just become more and more automated33:09and generally speaking i think the whole33:13idea of manufacturing even a lot of33:15coding as well33:16will become automated and commoditized33:19and so the value of the human the33:22systems engineer is going to be even33:24more so33:25on the earlier stage of systems33:27development right early on on the design33:29on the requirements on capturing the33:31intent33:31right not so much on the actual33:33manufacturing itself33:35um you know one important and this is33:37already happening now nvidia has33:38recently become33:39the largest semiconductor manufacturer33:42uh in the world33:43but they are a fabulous semiconductor33:45right and over time they were able to33:48grow at three times moore’s law right33:51and the reason they were able to grow so33:52fast and leapfrog over the competition33:55is because of the weight they put on33:58simulations emulations34:00and design upfront right now that’s34:03going to happen34:03it’s going to be just table stakes so34:06what happens in the early34:07phases of project is going to become34:09more and more critical34:11wow i didn’t know that about nvidia so34:13is where that competition was using34:16physical prototyping that nvidia was34:19eliminating much of that34:21into uh software simulation34:24that’s right yeah so um34:27and design and so when they were getting34:29um34:31wafers back then far fewer errors34:34than the competitors did it still took34:3618 months which is a general you know34:38moore’s law34:39that remarkably lasted for so long so34:42that time scale is pretty much set it’s34:44just set by the design the manufacturing34:47the testing it’s just 18 months but34:49because they were so34:50sure that the error rate was going to be34:52so low34:53they had three teams staggered six34:56months from each other34:57that when team one finished their 1834:59month cycle when they were third into it35:00the second team would start right so35:01they could release a new35:03chip every six months by having three35:05teams on 18 month cycles right35:07they were only able to do that because35:08they were so certain of their design35:12and now i think that level of35:15design rigor is going to happen in many35:18industries35:19around around the world many different35:21verticals35:23and i think uh consumers at least in35:25consumer products35:27are demanding that level of quality off35:29everything when35:31when i think when i was a child you35:33could buy something and it might not35:34quite work35:35and you might have to return it but now35:37it’s kind of unthinkable35:39if you buy a an iphone or something that35:42it doesn’t just work out of the box yeah35:46uh that’s absolutely true so the demands35:48of quality from consumers are35:49are much higher than they were before um35:52and i think it’s the demand of um35:56interactivity between even like big35:58giant35:59generators and when the big industrial36:01generators36:02you know whereas before they’d be uh a36:05black box and you turn it on and it36:07works36:07now uh the customers of these turbine36:11and power generators they’re demanding36:13access into how the internal workings of36:15the machine is right36:16how much power is it generating show me36:18the actual curve i want to control36:20something i want to turn it down at36:21night turn it up during the day36:23and and so forth um so36:27the manufacturers it’s not just quality36:29but they have to provide36:30insight and interactivity and the design36:33issue now36:34is much more important than ever it was36:35before36:36[Music]36:38yeah so i mean with all that change36:40going on what advice36:42giving people at early stage in a36:44systems engineering career36:46or people who are perhaps considering to36:48to do a career in systems engineering36:50um well i don’t um i36:53i think my uh advisor of undergraduate36:55days are kind of behind me now36:58but i think engineering generally and37:00systems engineering in particular i37:02think is still is a great field to get37:04into37:04uh it’s a difficult field but it’s very37:08important and continue to be and37:10continue to be important so37:12i would say number one i think it’s a37:14great career to go into37:15the world needs more systems engineers37:18than they have now there is going to be37:20a shortage uh37:22coming if it’s not already here and so37:24you know number one advice would be to37:26to do that um and number two37:29i would say for especially for37:33young people and recent graduates is to37:37um you know to really37:40focus on what it is they’re good at and37:43and take the work37:44uh seriously like i think it’s not37:47so important what you do is how you do37:49it right and really37:51look towards honing a craft right so37:53engineering37:54even though it’s you know it’s an37:55engineering discipline but there is37:56still a craft to it there’s still a37:58learning that goes into it and if they37:59have the mindset38:00of a craftsman that it’s a skill that38:02they just need to learn and own38:04that will serve that will serve them38:05well um as opposed to what title they38:08have what company they’re working for38:10oftentimes learning from a small company38:11could be better because you get much38:12more access than you would38:14you know at a big company for example38:17yes and38:18i often advise people great thing about38:21systems entering or38:22you know engineering in general mainly38:24you could essentially38:25any domain will needs you because as you38:28mentioned your software is eating the38:30world and38:31even if you wanted to run a hospital38:33like a systems engineering brain would38:35probably help but you run that hospital38:37better that like38:38these these skills can be applied um to38:41any38:42uh problem that’s uh i think really an38:45excellent38:46excellent point that is not made often38:48enough38:49you know what we’re seeing now is people38:51change careers uh much more often than38:53they have38:53right i’m a perfect example of that i38:55went from being a prof to a running38:57company that’s38:58that’s a pretty much a 90 degree turn um39:01and um and39:04just because you know training39:08as a systems engineer is prepares you39:11for39:12more than being a systems engineer it39:14prepares you for a whole39:16slew of tasks it’s kind of a way of oops39:20it’s kind of a way of thinking about the39:23world39:24right and analyzing problems and that’s39:26going to be the key thing39:27and everything’s become more complex a39:30systems engineering kind of39:32approach or an interaction kind of39:33approach is is a39:35key yes and you mentioned about how39:40the sorry in the future where the39:43generation of the39:45the artifact or the the solution is39:48potentially more automated but being39:50able to determine39:52what what is the intent as you put it or39:55what is39:56the um what are the requirements and39:59all of this is needed in any industry or40:03even a governmental level like um the40:06more people who are who are thinking40:07like this i think the better chance we40:09have in40:10in solving solving our big problems you40:12know globally40:13that’s right yeah i totally agree that40:15and we see that lack of40:19you know if you’re trained to think in40:20this kind of a40:22design level thinking let’s call it you40:24kind of see these problems all over the40:26place40:27and you can look at some of the40:28solutions that get rolled out40:30and it’s clear they haven’t thought of40:32it from a holistic point of view first40:34right they thought if they could40:35sequentially do all the components40:37then the whole system will work and it’s40:40almost40:41never the case given some sufficient40:43complexity that that is the case40:46so something that maybe was done a bit40:48more up front40:51analysis and upfront thinking would40:55um really produce a much better result41:00much quicker and then you know there was41:02a saying in astronomy that41:04the best way to build a 10 meter41:05telescope is to first build a 5 meter41:07telescope and then build a 10 meter41:08telescope right41:09and and that you know you want to start41:11small to figure out what all the41:13problems are and then41:14and and then go bigger right another way41:17to look at41:18to look at it is to think of what are41:20all the pieces and how do they interact41:21with each other right41:22and which ones should interact which you41:24shouldn’t interact and take care of that41:25first before you worry about the41:28what each individual component does41:31and so for your own career how have you41:33found the transition from that professor41:35into the41:35the ceo role where presumably you have41:39quite different responsibilities41:42i do and i don’t know how much of um41:49of my experience is due to my training41:51or due to my personality so some kind of41:53weird mixture mixture of two41:55but um i there are actually some things42:00that i find quite similar between doing42:03uh fundamental research in in42:07quantum theory and running a company42:10strange as it may seem the reason the42:13things42:13are similar is that you’re never sure if42:16what you’re doing is the right thing42:18and you’re always kind of worried that42:20you missed something or you screwed up42:21something something can come and bite42:22you so you’re always testing you’re42:23always thinking like how can i42:25make sure this is correct what can i do42:27to and you’re always kind of looking at42:28this one area and say okay i’m going to42:29fix42:30this and something else breaks and you42:31run over there and try to fix that42:33that’s kind of similar i would say for42:34running a company and and doing physics42:36research42:37they’re both complex systems and42:41and oftentimes the42:45you know neither the inputs nor the42:47outputs are welded the system itself is42:48not well defined right and42:50part of it is because you just don’t42:52know enough to define the system42:54part of it is that the system itself may42:56not be defined because there’s a kind of42:58a part of a creation42:59there you have to define the system that43:02you’re going to then diagnose and fill43:03so those parts are actually quite43:05quite similar the difference43:08is and there are significant differences43:10as well is that when an academic career43:14especially in physics i don’t want to43:15generalize because there is a lot of43:17diversity in academic circles but at43:18least at least in43:19physics that is really the main thing43:22is to get it right right there is a43:24system which is nature that you’re43:26trying to figure out43:26and your goal is to try to get it right43:28right um whereas43:31when you’re uh running a company43:34that’s not possible right uh and even if43:37it is possible it’s43:38always a bad idea right so there is this43:42idea that i’ve learned over the years43:43running a company that sometimes you43:44know you always kind of want to make the43:46right decision43:47but sometimes you have to make your43:48decision right a lot of it is very much43:50more subjective43:52there is no final arbiter of uh43:55it’s always trade-offs right and so it’s43:57it’s and you have to make a decision you43:59have to move forward44:00right even if you don’t know what the44:01answer is then pick one and just move44:04forward44:04and and try to make it work that’s some44:06of the differences44:08is the speed of decision making as44:09opposed to um44:11in academia yes and i guess with with44:15clients it’s like if you’re44:17just improving the situation even if you44:19have you’re not making it into the44:20perfect situation but improving the44:22situation44:23is in itself valuable uh44:26it is it is um and in particular for a44:30company like us however44:32we have to be a lot better than the44:34competition because44:35we not only have to to win over our44:37customers we have to show them a44:39solution it’s not only44:41incrementally better than what they’re44:42doing now because they have a switching44:43cost as well if it’s a large44:45organization44:46you know that they may say okay it may44:47be better but we’re not willing to44:49switch44:50just for that so we have to be a lot44:51better so you really have to re-envision44:53the problem and really44:54get to the root cause of the problems44:56we’re having and offer a solution that’s44:58very clearly44:59far better than what they’re doing now45:01right whether it be45:02far better in terms of the technology of45:05the deployment45:06of the user experience whatever vector45:08you want to measure used to be45:09it just used to be much better yeah i45:12sort of45:12yeah i wasn’t appreciating that if you45:14are a a smaller company45:17well one there if there is a switching45:18penalty but also yeah potentially45:21it can be for for that um the manager or45:25whoever makes the decision in the45:27the business maybe they have to justify45:29why they’ve45:30hired a smaller outfit as opposed to a45:32brand name that45:34their bosses may be more familiar with45:36that’s right it’s oftentimes a safer45:38status quo is always safer because45:42it is a status call after all right45:46so um you know given you’ve got these45:49big plans for um for45:51the next 10 years i mean are you hiring45:53at your your company45:55we sure are and uh and45:58i hope your millions of viewers will go46:01to qracorp.com46:03careers and have a look at what’s46:04available right now46:06we are hiring engineers software46:08engineers in particular46:09um and so you know we are looking for46:12them actively right now46:14we are looking for a product manager uh46:16to round out our team46:18um and that obviously those are critical46:20hires for us right now as we’re46:22looking to expand and always more46:25sales people always looking to hire more46:27sales people um46:28especially technical sales is is you46:31know46:31they’re not easy to find right um46:35i guess ideally you’d like them to have46:36a bit of experience either in a46:38particular46:39domain or in sort of selling46:43software engineering software to um46:46to systems engineering clients46:50yeah i would say all things being equal46:55we’d prefer to have more experience than46:56less experience46:58uh the issue is not only are all things47:00not equal it’s more like none of the47:02things are equal47:03so in the end we really want to get the47:05right person right and the way i think47:07of it now is if you think of say a47:08junior47:09engineer versus a senior engineer there47:11is47:12a broad distribution of quality among47:14junior engineers47:15there’s a broad distribution of quality47:17among senior engineers and there’s a big47:19overlap between them right47:20and so the important thing is to find47:22the right person who has the right47:23mindset47:24a growth mindset is important being able47:26to work at a small company where it’s47:27not just47:28you know here are your tasks for the day47:30and and go to it47:32we require people with a lot of agency47:35to help47:36develop the company and go forward and47:37really47:39build something new i think the47:40requirements engineering itself47:42is a relatively new field and so the47:45market itself is developing47:46as well as our product inside that47:48market is developing so we need people47:50who can think at these different47:51different levels47:53that’s great and i mean how have you47:55found the pandemic47:56um i don’t know what the situation is in47:58canada right now this is48:00june 2021 but um have you been able to48:03almost be business as usual given its48:06official work is more on computers48:09it’s definitely not business as usual48:11because the world is not as usual48:14but it had but i would say we have48:17adapted better than maybe other48:19companies were because48:20we’re a software company and we are48:24physically capable of working remotely48:27whereas if you’re actually building48:29something48:30you have to be in a lab and you know if48:32you’re on a restaurant that’s even worse48:34um so in that sense we do better than48:37both48:38you know if you look at what ha what is48:41the impact of48:42kovit generally and the rise of remote48:46work48:47so move to remote is obviously the48:48biggest uh48:50surface change that’s happened because48:52of covid48:54um the way i view it now48:57and it’s the view that’s evolving over48:58time it’s really quick covet has49:00accelerated49:01existing trends by five to ten years49:03right there already was a trend49:05towards remote we sort of leap frog over49:08it49:08right now we know now49:11we may have suspected before but now we49:13know that the core technology49:14video conferencing broadband internet49:17wi-fi everywhere49:18cell signal stability what else cloud49:21storage and sharing49:23the core tech is good enough today and49:25it’s rapidly getting better so49:27many companies are able to work remotely49:31this probably needs some work on49:32processes49:34they do need to get a little bit a49:35little bit uh better49:37we’re all still learning productivity i49:39think appears49:40not to be as bad as people were49:42originally worried about49:44but things like the cadence of meetings49:46and how to handle you know if you’re49:47remote49:48things tend to be a bit more49:50asynchronous or they ought to be more49:51asynchronous49:52and so how do you handle that as a49:54company that was used to being people49:55co-located49:57that’s something that has to you know49:58has to change and be um50:00uh an adapted right but it’s pretty50:03clear remote50:03is here to stay maybe not a hundred50:05percent but but uh50:06we’re certainly not going to go back to50:08the way things were ever us as a company50:10but50:11also us as the world i don’t think yeah50:14i mean i’m hoping for a bit of a hybrid50:16model i think that there are50:17there are benefits to both sides and um50:21i don’t know a year and a half now of50:23not having a physical meeting like i50:25kind of i’d like to have a workshop or50:27something50:27in person but uh yeah i think that’s a50:30good point50:30uh and so some things are better some50:32things are worse right and it’s just a50:34different thing so50:35overall it’s not better or worse it’s50:37just it’s just different right so50:39i think one thing that is better at50:41least for me and many other people50:43is there’s a big reduction in travel50:47so if you had a big company meeting with50:49departments that were spread across the50:51the country or the world50:53to have a big team meeting get to fly50:55there to co-locate50:56so a three or four hour meeting plus50:58lunch dinner breaks51:00travel hotels a three hour meeting could51:02take like two or even51:03sometimes three days to complete now51:07with zoom and google meet uh51:10a two or three hour meeting takes two or51:12three hours to complete right51:14so the meeting quality is almost51:16certainly less than it was before51:17there’s no question about that51:19but there’s a huge savings of time that51:21maybe that savings of time51:23you know counteracts the the degradation51:27of51:27quality of the meeting which is actually51:29improving right51:31things that i think are bad are things51:32like team morale culture and loyalty and51:35mental health51:36issues especially because you’re hidden51:38now if you’re51:39just seeing each other on a video51:40interface you know the social51:43clues to get other to help you may not51:45be there and so51:46that’s that’s an issue that needs to be51:48resolved i think yeah i think as well51:50for sort of51:50junior staff or if you’ve got you know51:53an established family and you know51:55a big house and home office51:59work from home can be quite comfortable52:01you get to see the children more and52:03but um i have heard speak i’ve known of52:07people who’ve um52:09their first job out of university has52:11been remote52:12and they get a laptop sent to them in52:14the post and52:15it’s like start your new job and um52:19it’s that i mean aside from all the52:21children and their sort of education52:23issues52:23um from from in school during the52:25pandemic um52:26i’m sure starting a job like that is52:30ideal um yeah that is uh52:33that is um i think a big a big52:36issue uh this feeling of belonging and52:39togetherness52:40in teams are very important i would say52:42you know in engineering52:44the teams that bond together and know52:46each other and have been together for a52:47long time52:48become better as a team just because52:51they’ve been together52:52a long time and there’s a compounding52:54effect and52:56when you’re remote and you join a team52:58remotely that’s just much harder to do53:01right so those are definitely some of53:02the53:03the negative things that we need to53:04still find some way to53:06to counteract that i think also a good53:09team it needs to have sort of failed53:11together and gone through hardship53:12together and53:14sort of helped each other through that53:15and um yeah but53:17that’s possible to replicate at least53:19with our existing electronic tools i53:21don’t know53:22so well it’s been fascinating jordan i53:24mean do you have anything else you’d53:26like to add to53:27to tell the listeners no i think we53:29covered a lot of ground53:31uh i appreciate it was fun talking about53:33i’m talking about this stuff53:34yeah yeah thank you very very enjoyable53:38so if people want to know more about53:40your company qra what what should they53:42do or if they want to53:43get in contact with yourself so53:46um our web address is qracorp.com53:51and that you’ll learn everything about53:53the company53:54if you want to contact me directly you53:55can just email me at jordan53:57qracorp.com um you can also find me on54:01linkedin on medium and all over the54:04interwebs54:05great great and yeah if you like the54:08sound of the company be sure to54:10uh check out the website and the careers54:12pages if you54:13you know like the idea of working54:14working um with qra54:17yes please uh or you know buy the54:20software and get it get it information54:21you’re coming so54:22thank you even better so thank you uh54:26jordan it’s been excellent54:27and look forward to uh seeing you uh54:30soon hopefully in the real world54:31if i can go thank you joshua yeah54:35or me in the uk yeah yeah great thank54:38you54:38okay take care joshua

Leave a Reply

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