UNICON - Low Code vs No code - What is the future?

[Music] it is uh it is now 2 10 uh central european standard time uh and we are delighted to have everyone here with us today for our panel session on low code versus no code what is the future we're delighted to be here with you today um just a few uh housekeeping reminders before uh we do some introductions for the panelists if you have any thoughts or questions please feel free to go to the session chat window and put your thoughts in there if you have any questions there's also a q a um section i believe we can use up in the in the uh in the sort of the chat box on the upper right hand side of your screen so please feel free to make use of that um we welcome everybody here today for our panel discussion on low code versus no code um i'm steve hilton i'm the moderator um i'm excited to have here with us today uh mustafa sakausas and deon dimich casey bison and michael hessenblas welcome guys thanks for being uh here with us today thank you for taking some time out of your mornings or your afternoons or evening to get started i'd like to give um each one of the panelists uh just a minute or so to introduce uh himself uh so that the folks listening to our uh panel discussion today sort of know our different perspectives uh casey why don't why don't we start with you good day uh my name is casey bison i am currently with uh with joyant a uh a samsung company my uh uh where i do product management uh consulting and uh my background is in uh sas and infrastructure as a service uh both both b to b and b to c uh and a really special interest in the usability and and how we interact with these systems but additionally at a background in how we operate our systems and and how we deliver them uh in production um eager to eager to hear where we go today with all of this thanks casey thanks for joining us deon thanks thanks steve my name is diane dimich i'm engineering manager currently working in volvo i'm passionate software engineer and serverless is kind of my domain which is closely related to local node code kind of topics i have personally developed some of these tools and systems and i will gladly here to share my opinions and views on the future thanks welcome and thanks for joining us dianne michael all right um yeah thanks for having me my name is michael hausenblus i'm a solution engineering lead at aws web services in a service team so the folks who you know write and operate the services that you many of us are using on a daily basis um i've been spending the most couple of years in containers like quinnipies and messes and stuff like that before that i was a data engineer and before that some 10 years in research applied research so the educational aspect of that is always something very close to my heart excellent thanks for joining us today michael and uh mustafa yeah thanks steve thanks for having me too uh originally i'm a computer scientist uh but uh also i'm a serial entrepreneur and y combinator alumni uh i am we are developing a no-code platform which helps you build customer-facing applications without coding or if you want to you can also code but so i wanted to combine all my uh background experience both on the engineering and entrepreneur so we are mostly focusing on startups who want to build their sales products excellent mustafa thanks for joining us and just so folks know my name is steve hilton uh i'm with a company called mac nation mac nation is an iot test lab so we test um in a lab environment iot software sort of the platform stack in the software stack and mostly what we end up doing is iot performance and scalability testing we built a piece of software that actually simulates iot devices at extremely large scale it's a piece of serverless software that we built actually on top of aws services and we use that to simulate iot devices at massive scale and help enterprises do performance and scalability testing so thanks again everybody for for being with us here today for our panel discussion on low code versus no code what's the future um so once again feel free to throw in um your chat or your questions if you have any um i think to start maybe um let me give a few definitions um we've got these two phrases low code and no code so i thought i would go to this the font of all information wikipedia and um definitions uh for what a first let me start with a low code development platform a low code development platform provides a development environment used to create application software through a graphical user interface instead of traditional hand-coded computer programming okay so once again it provides a development environment used to create software through a through a gui right instead of traditional hand-coded computer programming and that's low code now no code is defined on wikipedia as allowing programmers and non-programmers to create application software through a gui and configuration instead of traditional computer programming so once again the no code development platform is allowing programmers and non-programmers to create application software through a gui and configuration instead of traditional computer programming okay so we have these definitions of low code and no code um i think maybe i'd like to start off asking our panel um maybe um let's start uh with michael who who do you think uses low code or no code or both right in my experience there are two um distinct classes or two categories where i've seen that the uptake and interest quite big on the one hand you have this educational environment think of mit scratch where you know especially um either younger folks like you know with kids teenager range and they might look at that as a nice entry-level way to you know get into into the world of programming um having a fast feedback loop where you know they either enter in the computer or with some you know iot related stuff some sensors or whatever relatively quickly get something built and see that it work you know in nitty-gritty details having to learn what is a for loop you just plug something in and see you know repeat uh for a certain number of times so that's the kind of educational um class the other is um i'm thinking for example of what we have with amazon honeycomb um essentially enabling folks in in slightly um you know related areas who are not you know coders per se but they have you know their bi folks or data scientists or whatever who not necessarily want to or can uh really code per se but really have you know some problem solving um capabilities and but they want to deliver that kind of experience it could be a query it could be you know dashboard or whatever um in an effective way so that's the kind of like folks who have a a somewhat related activity but not necessarily their main job or maintenance or whatever it is to actually drop down on the on the low level program that makes sense agree disagree folks other types of folks who use this stuff uh i can share my opinion here uh since i kind of like more more abstract version of the low code versus no code and um it's more about that we allow people without coding abilities to create applications and processes and if we go i think every one of us uh today are using this kind of uh capabilities through other applications that allow us with a small kind of investment in drag and dropping to adjust these applications to our needs and if i want to go even to amaze you a bit if you ever use something on aws you can imagine that it's a drag-and-drop configuration for your infrastructure if you want to send a email from microsoft office world it's also a low code kind of investment for mail merge and sending emails so today i think in our time we all are exposed to some kind of low code engagement by definition that we find on marketing on wikipedia which is most marketing driven we design like more on application building using visual tools but i come from amazon developing alexa for two years and using a voice as an input mechanism is also a law called kind of engagement so we should think it more abstract if we want to see what the future looks like fair enough fair enough casey i think you were going to say something yeah i really want to double down on i think what both deon and michael have said in terms of the audiences that are using these um i think the educational piece is incredibly strong and the way people are learning from it but when i see generally business teams functional teams using these tools to solve their problems because they're underserved by their i.t services because they're underserved by the tools that they have and they build solutions for them using air table or or similar things that that are honestly very sophisticated applications that that they they build to solve their specific problems in a way that outsiders are unable to solve or the systems they have are unable to solve and that's really empowering um and i think it really speaks to the way technology entered the workplace uh over time especially especially pcs entering the workplace from the bottom up rather than big uh big uh larger scale computing from the top down so interesting stuff going on there yeah mustafa you can you can add to this question also um with some of your thoughts or i also wanted to get sort of your perspective on is there a difference between low code and no code or frankly is it just sort of like a marketing buzz thing and and they mean the same thing or yeah so if we create the spectrum and put the two edges in the logo then no code so we see a trend that both systems are getting into the middle uh so so there are local tools but you can also do many things without coding in those platforms and also there are no code tools but you can still do some coding inside it because uh all the platform owners want to make the platforms very usable and very easy so they have a no-code first approach and mostly but coding is still the most flexible language so far so you cannot convert all those power into visual tools so you still need the coding uh in some areas and if you want if you want to make your users more flexible you need to keep that doors open uh for everyone so i i think it's a term for today but in the near future i think it will not be the case that we will not discuss the no-code local tools actually we also see that gartner is uh categorizes no code under a subtopic under low code and and so in my opinion there's not much so much thing but on the marketing perspective if we are talking about enterprise id they all talk about low code but if you go to startups freelancers and the bottom up actually approach you usually use the no code turns out it's kind of marketing term but it will actually converge in the near future fair enough thanks thanks thanks mustafa um so when when we think about sort of the use of low code or no code um solutions why do companies use low code no code is it driven by a technology need is it driven by a business requirement does anyone have any thoughts on that what's really driving folks to that to use this stuff yeah casey why don't you go first yeah i'll i'll jump in because it's something that i'm really passionate about i i think i think anybody makes the top down decision like we're going to use no cow no code or i don't think many people are doing that i think it's this is really driven increasingly or not increasingly this is really driven by by people solving their own problems people who are in these business areas seeing a problem that they need a solution to and and building a solution that works uh for it and that's that that empowerment of people i think is is really a strong motivation that people are seeing across across business um where where they are realizing that they can empower their teams by letting this happen and so i i think it is in a lot of ways revolutionary revolutionizing i.t in that way diane and then michael okay so uh if we take a look on the nature the the the the systems that are adapt fastly will survive and this the same for the businesses as well and we see that from the day one when these new technology aka computers were introduced and this is not a new thing uh it started with definition of new programming languages uh that afford generation that allow us to work on a more higher uh aspects of of programming now it's kind of on the steroids then everybody can work together interacting with different solutions that are provided by many yeah if i just want to to say that using this kind of drive that businesses have and us personally as well to be more efficient move us in this acquiring more customers customers are acquired by letting them following their problems and this kind of enablement give this law code a boost that we see now but i want to say it was there from the beginning yeah fair enough fair enough michael i think you were going to say something did you want to add yeah yeah i was actually heading in the same directions in there in terms of you know if you look back some 20 30 years ago and you have this the rise of the 4g languages and you know um it's kind of similar discussion like you know why would i want to use a high-level program language a 4g language rather than you know assembler for example and turns out that you know none of those things ever goes away there is still a you know sizeable amount of people who develop you know assembler and there are certain use cases where you need to drop down on that level might might be because of the constructed environment you might have or performance or whatever and the same is happening now right you have new areas you have no code local but you also have machine learning that you know may or may not replace us over the next 50 years that doesn't mean that the other things that we used to do uh go away there are certain things like i don't think that we have a lot of um you know punching cards anymore but you know most of those protection levels uh stick around for for good and you always need some of those experts to cover that but the majority if we talk about mainstream if you know we talk about i don't know 2050 and you're looking at what is the mainstream way to create and deploy applications my bet would also be on a combination of low code no code and um machine learning assisted you know creating certain algorithms or parts of the application good deal good deal what i'm realizing none of us have said yet is that this is substantially increasing the number of people who are building applications and that's a really key part of this we're i think it's opening the door to even more people um building building the solutions that they need and that's that really needs to be said here yeah exactly actually i totally agree on casey so and these local no code platforms also create a carrier carrier opportunity for most for many people for example uh yeah actually developing software really pays good but not everybody can do it very well but a low cost no code platforms are a very good step for them to start on and so that's that's one of the reasons that uh pushes the trend but also i agree on everybody that business drives this trend because yes there are so many uh requirements and also and the idea is getting underpressed with so many uh things so they wanted to do these things very fast uh so they wanted to create a solution and they wanted to respond back and this digital transformation also pushes uh forward so these are the trends that make it and that's very helpful and very insightful it's also a good segue into my next question for you guys um on this topic i was going to ask about the pros and cons of low code and code but honestly this conversation so far has sort of illuminated so many of the pros of it it sounds like we're a mutual admiration society here and we are all completely behind it why don't we start with the cons what's the downside of low code no code since we've talked about so many of the pros already thoughts casey i saw you doing one of these things i didn't want to be the first one to jump on it but now you um i mean so really clearly the the cons here are we have shadow i.t being built all over the place with with what's going on here we have something where increasingly especially the way this is this is entering from the bottom up we have knowledge and understanding of of increasingly sophisticated or complex business processes that can be lost when people leave and so some very classic i.t management concerns that are really hard to address because with with this now uh because these applications are really really built by individuals for their individual work processes and not generally managed in the same way that centralized centralized i.t is let's say you want to let's say you need to do qa on these things how do you who's responsible for that how do you how do you manage uh continuous delivery for these things and these platforms generally don't have that sort of thing you make a change that changes live immediately you can't really see and test the different versions of that in the same way you would and maybe reverting being able to get back to a previous version of something or even compare versions of something could be impossible in a lot of these platforms so there's a huge for people approaching this from an engineering perspective where you expect to have things versioned you expect to be able to do continuous delivery with with uh with a lot of testing around this stuff you run into a lot of barriers um from the it perspective as i say uh that that side is is a huge problem so there's a lot to a lot for us to dig into here as a group i think but i'll stick on those mustafa do you have any thoughts about uh some of the downsides of low code no code yeah absolutely so yeah these tools actually creates an abstraction on top of coding and sometimes they create a very high level of abstraction so they actually make everybody very away from the coding so these are done this has some downsides because if your application uh gets larger and more sophisticated then your platform doesn't answer your needs so sometimes making so much abstraction and creating in a very new language for the people it creates a problem when things gets complicated so this is one of the downsides so on code 2 actually we are against the stop so we are not trying to abstract that much uh so we still you want to still uh keep the doors open uh and the other is uh you need to choose your platform very carefully something that you need to get vendor locked in and so if the platform has problems such as scalability and other things and those are the other uh downs of the platforms good deal hey before you the rest of you guys jump in on some cons i want to remind the audience if you have any questions go under chat and session which should be up on the right hand side of your screen and feel free to put your questions in there we'll get to them shortly guys other sort of downsides of low code no code yeah um so it comes to mind and without you know trying to be a gatekeeper and and discouraging folks to you know try try things out to um you know have have an impact there um if you consider software engineering which you know as a discipline you learn certain things in terms of um in case you pointed it out in terms of quality you know testing et cetera et cetera and if you move into certain areas there's also liability right like um i if i'm entering a plane then i would like to know that the software developed there um was not put together in five minutes by someone who you know learned a local tool yesterday so this is not this is not to say that you can't do it um but you either need to make sure that the tool itself embeds that has this kind of like um you know checking and then making sure that the bar is met uh or exceeded uh uh built in or um you acknowledge the fact that it's it's relevant for certain phases think about like retool for example you might do that in prototyping right but then if you are actually developing a ui you might want to pull in ux designers you might want to think about what are the important things how not to clutter ui whatever but you know retail gives you this moving fast right you don't need to spend three weeks to put together a ui to see how that would work so there's certainly this challenge there and again i'm not trying to gatekeep or whatever here but we should be aware of the implications in terms of the the quality and the bar there okay so let me be a devil's advocate here i don't see any downsides of the load code approach whatsoever i think whatever we mentioned here was actually not the downside of the low code approach but maybe about us not understanding the requirements needs processes and the quality gatekeeping as you all mentioned here so if we have a like a one coffee it would be easy to select purple one if you have a bunch select from all kind of consequences can emerge about different qualities different taste are we drinking the appropriate coffee inappropriate time or tea even worse so i think we are we should somehow be aware that the number of people who can actually build something will elevate the complexity of the environment but will not somehow remove the all issues that we have with our understanding the environment and i think our goal is more to as mikhail said educate people about the environment and then as there is a a well good saying with a huge power comes the huge responsibility so be an owner of what you are using and just go for it there is no other kind of perspective um so fair enough i think that's that's my take at least no good that i appreciate it you have some really good perspectives on some of the downsides of a low code no code from panelists here we've talked about a lot of the pros already the the positives of using low code and no code i want to give um some time for the um audience to ans ask some questions also we had one question from david about how do you feel low code will change or transform the typical software application development cycle okay so you feel that local change the software application development cycle anyone have thoughts on that so go ahead and stuff please yeah i can next you can go okay so so first of all i think the the applications will be more granular and it will be more diverse and more specialized so the development cycle will be shorter but more straight to the point and more oriented to integrations with other systems yeah good point mustafa did you want to add comment to that as well yeah so if you think about the software development cycle so sometimes you start with designing then you start developing then you're deploying them in maintaining uh in local no-code platforms actually this transition is very smooth for example you can design in the same tool you can develop in the same tool you can deploy in the same tool you can maintain in the same tool so the whole company can use the same tool and this transition is very smooth for everybody and you can even manage your project in the same tool so this is an advantage so this will i think change the whole uh life cycle but also the software stack is getting very complex nowadays so you have a software stack and when developing so you have an infrastructure site so you have different teams different uh technologies so combining all this into one and needs a very big team sometime and also maintaining those stuff needs another expertise uh from very uh highly paying talents so local local platforms also takes care of uh those issues too and for everybody and and it will also make change on the life cycle of software good good deal thanks for that i'm gonna i know that everyone would like to comment on this but we've got some other questions coming in so let me roll on michael you have 45 seconds one of the questions was what are some of the aws solutions for building low code apps this is your 45 second advertisement for aws if you'd like um i can just encourage you to check out uh aws.amazon.com there are way too many services that i i'm familiar with i the one that i've been using myself is honeycomb um but there are many many and you should not think of the services themselves as like monoliths they are services that have certain aspects that are that fall into that category um and to me the most important or interesting thing are machine learning related things that you can use to boost or accelerate or to um shortcut certain things that you know otherwise take a lot of of manual labor and manual coding thanks michael appreciate it we've got an interesting question from sander do we are we missing something like git for low code that would be equally easy to use as the low co2 tools themselves casey did you want to start on that when i think about the way uh low code really does and it's all of us i think have said something about how low code does change the software uh development life cycle in some way and that's a big thing the you know we're going through another generation of learning in terms of how to successfully build software that works that works reliably and and so the lessons that we learned in doing that in in our traditional environments we're going to have to relearn those in these environments and when i talk about the lack of versioning that's earlier that's that's exactly what i had in mind like there's there's very limited ways to inspect previous versions see differences um and and used the techniques that we've developed already for this in these environments and that's that's something that we really need to think about right can i quickly jump on that because i at the same time agree and this degree because on the one hand if you look around you know if you have i don't know google docs or whatever like you never ever explicitly save it's an autosave all the time right you you don't really worry or care about that the other hand like this this explicit thing like you know i want to jump to this version you may or may not be able to do that so making that more explicit sure but that's that should be a general um you know feature that you can drop down one level and then you know the people who know what that means can deal with that but i would argue that by and large like all of these systems do the autosave with that you already have this at least implicit version all good points all very good points uh let's see we had a question uh um from mark around a panel any thoughts over data security in in in mark's experience and he says he's a fan um he sees that these apps they often do not get trapped within the same security testing that different apps would be subject to so the low code apps may not uh be subject to the sort of security testing deon did you were you going to say something yeah yeah yeah so as we discussed before about it's more about our understanding of the processes and requirements and security is one of them and there must be enough knowledge in organization or individuals when they are using uh third-party tools or lock code or no code kind what kind of security implications that involves and also privacy implications as well but this is not different than what we are using now and i can say that even today the most rigid companies has moved to the cloud already so we get this kind of knowledge enough that we can work with the with the outsource vendors sharing some of common information around and taking care of the security the more interesting topic for me will be like vendor lock-in which is also usually uh bound to this one as well and i think law code with this diversity give us even more to choose from and easy to switch if we desire thanks deanna any other thoughts about michael michael about security i tend to agree uh with what david said um and and knowing that you know at least you know in my case with aws we have this shared responsibility model where it's clear where we draw the line uh this is what the customer is responsible for this is what we are responsible for um i just like i'm not a lawyer but i can imagine uh an issue where uh in terms of liability if it's not clear if i'm using a no code local tool and it's not 100 clear that it's actually me who's responsible for you know things like in europe gdpr compliance etc etc where i kind of like in the same way that um currently i can't go to twitter and say like hey you're responsible for the content you put out there um is it my responsibility to you know be gbr compliant or can i you know make the the tool provider whoever provides the environment responsible liable for that so these are things and i agree this is not like unsolvable or whatever all the things that we raised here that you call downsides i would just call challenges those are like 50 to 100 startups ideas so whoever is listening out there pick any of those solve it and you will be the next you know whatever i i wasn't sure if anyone else was going to uh i had a comment on the security question i'm gonna i'm gonna i was i was debating about joining that silence because michael's bringing up of gdpr uh especially highlight here right to be forgotten and anytime anytime you're building uh something that has data flowing through it and and we're talking about something that has really distributed the creation of systems that are managing a lot of data who's properly ensuring compliance for right to be forgotten is going to be a really big thing i think in a lot of these systems and and that ccpa is is uh also adopting it also has similar provisions so uh so it's a it's an increasingly global concern but again these are things where i think we've we're developing practices in traditional systems that we have to figure out how to apply in increasingly distributed uh systems that that it's not new and and it is however a new generation of solving similar problems that we really have to figure out very very good point you know as we're sort of approaching toward the end of the conversation let's let's hearken back to sort of the beginning when i when i said the title of this um conversation say low code versus no code what is the future so let me ask a sort of a wrap up question for our panelists around the future is low code no code enough enough to give an enterprise what it needs in the creation of applications or frankly is low code no code simply an intermediary step an intermediate step toward sort of the enterprise in the future finding what it really needs so let's think a little bit about that future um world of the enterprise what do you guys think so let me jump straight away to it oh mustafa i always kind of come before you the low code will be majority of what a lean business model will be in the future but there'll be only small parts that are really core business for any enterprise to keep and using the capabilities of others will become a must-have approach and if you think about it what is the the best case scenario for any activity that we have is to have an algorithm aka a cookbook that will solve our problem and walk us from point a to point b what a great chef do is put a little spice and enterprise will keep that spice everything else will be in the cookbook mustafa can you can yeah some additional spice there uh yeah actually i agree on them and also yeah in local knockout platforms what you are building is more context aware than developing coding from scratch so if you want to put ai on development so it is much easier if you start with low code no code tools because uh because the system knows what you are trying to do such as putting a data grid or accessing some integrations and other things and also it can create more assistive uh things uh on local network platforms so uh it will be easier it will it will be a middle step but i also believe that uh these no code local platforms actually will be a platform for developing components and applications so it will not be like developing things but it will mostly like uh putting building blocks uh together connecting each other so these platforms makes it uh available for everybody so it it is just a middle step for now but everything will be built on these platforms uh in the future casey michael thoughts about sort of the future is this an intermediary step that we'll sort of see or not really i'll jump in that uh and and i'll join uh i'll join what was said i think here but but let me rephrase it i think sas is the future uh and and integrations between sas applications which will increasingly look more like some more sophisticated more modern versions of mail merge uh that was brought up at the beginning of the the conversation i think where where people are are integrating between systems that are that are maintained and provided by others and we are doing small bits of logic for how we integrate those systems and and the power of that is huge um but uh there will be some aspects of low code or no code in there but for the most part when people feel as though they need to build something that is truly new and different i think it'll start out prototyped in in sas and then and then as they prove it go into in larger environments more more sophisticated environments nice insight casey mike michael where are you going to add something yeah i'll go off it mentally different tension that is i think the unspoken assumption is that we will continue down the path on you know silicon and for neumann architecture and all of that if you know in five years time or next year or whatever uh quantum computing or biological computing or whatever comes along so all these paradigms are essentially you know out of the door they're like you know legacy essentially so under this linear assumptions of the non-clinton christian disruption if you wish the the more like traditional innovation uh sustainable innovation um sure what we said makes absolute sense i do see the machine learning aspect as a as a quite important building block to to you know automate certain things and to to um make sure that that you know quality is met testing all the things that we discussed today um but yeah i i see a bright future there absolutely i'm pretty bullish and pretty updated around that excellent excellent anyone have any any uh student clue i just want to somehow complement the whole conversation about the future uh by saying that in order to survive we need to better understand our environment and our environment is complex and as we unveil parts of it and become more and more complex for us individually or as an organization we cannot focus on the whole environment so we have to divide that and there are some specific organization that will uh be focused on one aspect of that environment and for us to leverage we have to communicate and integrate with these other participants in environment symbiosis will be the only way that we can survive so for that reason this kind of approach that enable us to easily integrate either through sas models or using tools that are low code regardless of way how we will put inputs in is the future how we will continue to survive i i i think this concept of symbiosis is really important and it has come out loud and clear today um as you've just uh brought up deon the idea that there will be a series of tools sas and serverless and infrastructure as code and all kinds of different tools and development philosophies and development models and there will be as we've seen uh for so many years sort of this symbiosis and this sort of joining together um one of you guys uh casey actually brought up this concept of sort of the integration of systems is so important today and i think you've rightly brought that up um gentlemen i want to thank you i want to thank you for your time today i want to thank you for your wonderful thoughts and insights um uh audience thank you also for your great questions um and sort of the running dialogue on the side we've tried to keep up with that as we've had the conversation today um once again a special thank you also to waylay uh for uh hosting uh unicon today and all of the panel sessions thank you so much thank you guys thank you for having us

Channel: waylay

Views: 607

Keywords:

Id: l3G_iRCmuuQ

Length: 44min 51sec (2691 seconds)

Published: Fri Jun 25 2021