ServiceNow’s Jithin Bhasker on Generative AI and Low Code App Development

hi I'm James Maguire and on today's easts we're talking about how the combination of low code software and generative AI will dramatically change the process of software development probably in some pretty unpredictable ways to discuss that I'm joined by jithan baser general manager and vice president for the app engine business at service now jithan very good to have you with us today thank you James great to have great to be here you know I've got some in insightful questions about generative Ai and and software development I I know I'd like to get a sense you know the short sense of what service now does of course many people know it it's an extremely well-known company but for those who aren't as familiar with service now can you give us kind of the the 10 to 15 second version of of what the company does absolutely Well Service now is the market leader in Enterprise workflows and the way you can think about it is we cover the entire Spectrum of an Enterprise company all the way from your front office which could be customer service management to Mid office which could be anything to to do with the you know Global business services and the back office which also includes finance and supply chain in summary we kind of cover the whole spectrum of customer service it um finance and supply chain and also we have an amazing platform called service now platform which is the one which I lead and own at the moment and which just a protocol app in sounds like the kind of job that would keep you busy to be sure um all right so so so lood and and generative AI That's that is a a really big deal this this year obviously ever since chat GPT came on the scene last November seems like generative AI is the only thing that people are talking about what what sort of Trends are are driving these two technologies here at late 2023 what do you see out there in the field yes for sure I know it's a it's a big uh bus word in the market in terms of generative AI uh well um there is a powerful combination between the generative Ai and the low code Technologies and I'll I'll kind of take a step back and give you a little background in terms of how those two are emerging right so um you know some of the challenges and the trends we see in the market which is actually going to accelerate the adoption of these two technologies together is number one the there is a NeverEnding backlog of requests for it to be able to build a new application automate something or build me a new dashboard or analytics or any of those that you can think of and often what happens is that it has a certain limited set of you know skilled talented engineers and developers but above and beyond finding the people who have certain skill set in different platforms and tool sets it's also extremely difficult as an example you know uh there are 400,000 jobs to be created into the software development world over the next 10 years or so and we have a huge shortage of the number of people who are really skilled in kind of ramping up into the cve of that that role the skill set and all of it together so what fundamentally the combination of two is where there's a huge dependency on Talent huge dependency on those Talent actually knowing these platform specific platforms and tools and and the Technologies and at the same time there is an immense amount of this whole digitization and digital transformation creating a ton of backlog for it on a consistent basis according to Economist the the reset says on an average most of the companies have six to 12 months of a backlog in terms of the request which is actually coming through that's where the combination of generative Ai and the low code truly enables the whole Enterprise company you you you're talking about you know um team members who could be a part of Marketing sales HR or business system business analyst kind of a group who has now the ability to be able to build on certain local platform which could then then effectively reduce the need for it to be in bit so effectively the the way this whole technology is going to evolve is generative AI will reduce the barrier for anyone to be able to get start to build a certain automation or an application and then you know the way it will work is when you write a simple English statement which says Bild me and marketing events tracker the generative AI Technologies especially on now platform um which we are you know launching a variety of capabilities like text to code text to flow text to app is coming in the future more or less this this technology will write 80% of that application or automation for you now the last 20% or the last mile is where you're really customizing now you don't want to code you don't want to script you need to have certain easy to I would say configure click and drag and drop and that's where the low code platform really comes together so that's what makes it really powerful the combination of ability to build something quick then build it the final touches of the last % with the low codee you know Technologies which will help you actually deliver some of these automations fast interesting it seems like it's it relates to this big Mega Trend in the industry which is the idea of democratizing access and that there's probably some mid-level sales rep out there who can handle a low code application low code software application with the help of generative AI so you may not even need an expert developer to create this if if you have a low code you know software building system and generative AI correct absolutely yeah and we see hundreds of use cases which our customers are starting to build all the way from marketing to you know you're even talking about doctors in in the Forefront of the patients actually starting to really innovate on things like how can I get uh uh certain insights about certain discussion I had with the patient to be able to fully fully accessible across my team of you know uh team in the hospital or in my clinic where how can I actually put a workflow together we see some of these really amazing Innovations being built by customer doesn't matter what your skill set is it could be you know you've never had uh technical degree or a background or you had some sort of an inclination over the technology like a business analyst probably we are enabling all of those Spectrum where low skill to Mid skill to high skill personas and to be able to automate and build applications on the platform well with with that in mind I I I have to wonder maybe there's some young man or woman out there who's thinking you know their college age is thinking well I'm going to go into software development and and then they look out at the market and they realize oh generative AI is writing a letter of the code gosh I shouldn't even become a software developer and yet you're talking about a shortage of developers so how do how do you reconcile the these two concepts yeah I would say the gender AI has actually created a new demand for a new set of Highly skilled talented Engineers which you could think about you know there is a brand new role which is emerging in the market called prompt Engineers effectively right where you really need to and and it's a very very unique kind of a thinking you need to apply it's not only technology kind of a understanding you need to have but you also need to really think through from the actual user or the industry or the actual customer in terms of what they intend to be because a lot often these Technologies right generative is all about how do you accurately uh infer or understand the intent of what that English statement or a prom could look like and it's a very different skill set Al together that's one set of you know high demand which we start to see emerging in the market the second one is like you know the all of this works on language models which are actually powering up all of the Technologies in the front end effectively you need to have a lot of Highly skilled data scientist and Engineers who could effectively look at all these language models continue to refine and fine-tune so that it delivers you you know highly accurate responses and and output the word hallucination has become a lot more popular nowadays and how do you reduce the hallucination of some of these models through this fine-tuning and refining these models these are some of the roles which existed in a in a in a minor way in organizations as a data scientist probably in the past but now that that has become mainstream so if you're someone going into computer science right now uh you are I mean you can definitely think about this highly Advanced roles which which is growing you know tremendously in the market every company is looking for someone who could effectively fine-tune the models and the fundamental demand of this fine-tuning also come from the fact that um you have a lot of Open Source models which exist in the market but more than that the best acur responses you get when you have real custombuilt llms uh you know within within the use cases of the industry or the customer base you are serving as an example service now is fully investing in building our own custom llms so that you have you our customers can take advantage of high accuracy rates in terms of the response they receive versus connecting with a lot of Alternate open- Source models in the market H interesting well all right so there there's a lot of possibility for for generative Ai and low code no doubt about it but there's also some challenges involved I mean everything doesn't go great so so what do you see some of the challenges I mean compliance bias lack of skills what's what's holding it back or what are some inherent difficulties yeah I think you know uh the industry has spoken a lot and enough about the biases which we need to really consider within the models well I will not touch upon the topic because there's a lot of literature out there but sure one topic which is really close to my heart while I'm you know leading this whole product strategy from a low code and how do you integrate the generative AI within is what we are now doing is we are unleashing this highly powerful capability of the combination of low code and the generative AI together for now anyone like the sales you know member example you spoke about earlier anyone can now build any of those applications on this platform now the the actual value of embracing an Enterprise grade platform like a service now could be is it's built in a way that you can get the right controls and governance in terms of who can access what data even though it is leverage through generative AI but at the same time be able to make sure that before you publish it goes to certain Security checks privacy check data checks and all of it so that you manage it as a stream work through a value stream so that you have the right governance and controls and guardrails put in together so our local platform especially the app engine we call it app engine management center effectively is built with the notion that you have the ability to fully control who is building what and what kind of a data access you need to have and at some point of time you need to manage the life cycle of of that application or automation which you built would you like to maintain it retire it all of those could be easily taken care through this value stream thinking which we have applied across through our platform at the moment through something called app engine management center so going back to your challenge what you said fundamentally I feel this is a technology we are unleashing to anybody in the Enterprise uh you know in the companies so you need to have the right guard rails and controls and governance in place so that it's not an apps spra you're you know you don't have a data I would say distribution issue or a control issue fundamentally from a privacy point of view it feels like a company will obviously very much need that guardrail process that's not going to be optional they will need that it's it also seems like um these generative Ai and low code software they they offer this whole possibilities but now we do need this additional sort of personal or company infrastructure in place the guard rail process so it's it does set up maybe some more staffers more concerns some some more worries for the the company that this is so easy to create as you're saying that's right I think that's what I think it's very important that while generative AI is a technology every company want to embrace and get started with at the same time making sure you're building that on the right Platforms in the market with the right guard rails and and and controls in place so that you're not only enabling the velocity of how many people can actually build application but at the same time us a CIO uh who who has a peace of mind in making sure that this is all managed a centrally through something like an igon Management Center which I actually explained about earlier so it's a right platform um along with the enabling the generative AI on top of it would make Wonder in terms of you not only enable velocity but at the same time you have the right controls on the other side well let's drill down to what exactly service now is doing to to serve this Market you you've talked about it certainly but let's make sure we cover that how how is how is service now serving this Market yeah we we you know um we've been investing in generative AI Technologies for the last many years and if you go back and look at the history of Acquisitions we have done uh this is in the bus word we just caught like a year back it's been an investment uh we've been consistently acquiring this highly Advanced AI Focus companies in many many years and um not only that like it's been laser focused on uh all the way from I would say uh we introduced our virtual agent capability in itsm a number of years back which is already powered with a certain amount of intelligent um Automation and intelligence at the back end through variety of this models which we spoke about but as we speak we already have capabilities like a text de code uh text decode is very focused on high skill developer who is building on a service now platform where you can simply write English statement on your IDE and more or less the the our text code generative AI capability will will build a whole whole lines of scripts and code for you so effectively you know uh almost your your your 2x or 3x faster in terms of being able to build on our service now platform so text Cod is something you know which we have launched at the market we have a lot of customers really excited to be today about it in addition to it we have also added number of capabilities from case summarization uh for if you're a customer service agent who is in the Forefront uh serving your customer you have the ability to provide our generative AI capabilities will provide the agent with the case summary of what the last few discussions they have had or their team had with the customers so that they they have a quick snapshot view in the past they have to scroll through all the case notes all of those different interactions and engagements and get to that core of it but in a matter of seconds our case summarization capability which comes with itsm or CSM which actually helps the agent to look get that summary level information in a matter of few seconds we also have number of other capabilities like uh next best action you're building a workflow or a or a process flow on our platform effectively our generative AI will automatically recommend or suggest I see that you built a certain type of a action now here are the few other actions which we recommend you add on as an example I received an incident from my it uh Network or the infrastructure now I want to be I want to be able to automatically uh notify someone on the development or it side as in when the cases actually land so we've added a number of this capabilities like all the way from a virtual agent to case summarization to next best action to text a code which has been phenomenal in terms of some of the creativity our uh users have been applying we have a lot more coming in the future and uh we are going to continuously release it like a monthly or a quarterly basis going forward all the way from you know text to uh app to text to process we have a number of those coming in in in terms of our Innovations very shortly it's an investment we are heavily making to be able to actually truly deliver that Enterprise grade uh generative a capabilities for our customers so clearly there's there's a lot going on and there's really a lot going on throughout the industry which creates the question uh you know where is it all going the future of generative Ai and and low code in particular certainly it's going to have a big impact I mean if we have this conversation in I don't know 2025 2026 what will we be talking about then where is this going say two three years out yeah great question uh James like uh the I was reading a prediction or a research report by IDC according to them by 2025 70050 million new apps will be built using locco Technologies 750 million correct yeah W and and and in addition to it the the many other research which also shows like um almost 75% of those applications will be built using low code platforms and Technology now add generative AI in the mix of the combination of low code and generative AI my hunch is it's going to at least double in terms of the efficiency and the velocity these teams are going to gain because as I was saying at the beginning of the call you've reduced the barrier tremendously for you to actually from an idea to a a build to a product to publishing it in the actual production has shortened so much because in the past when you have to start building an application you need to think about what the design would look like which is still relevant you need to have a certain amount of idea of design but then you need to think about how you actually set up the system architecture infra all of those components need to come together for you to be able to so you're reducing all of the time when you have a low code platform which is powered with things like a text to flow or a text to code or a text to app in the future where you will be able to build a app with the speed of your imagination in few minutes 80% the app almost 80% of the app will be built right in front of you by the generative AI capabilities with platforms like service now app engine and loow cod tools now the rest 20% you want to fine-tune the the color of the UI or you want to integrate um you know have or you want to add a new form or a field or or you want to change the header all of those could be now easily done using this highly configurable low code uh Platforms in the beginning so 80% would be done by generative AI in few minutes that'll be really fast the rest is where you really want to refine and polish the actual experience so my hunch or my prediction is the number whatever they have predicted will double because you made it so easy for anyone to be able to start building applications and automations now that that is really amazing that that 750 million figure and I love your phrase building an app at the speed of imagination I think one of one of the better phrases I've heard recently I'm going to I can't top that one um Jan I think you said it uh I've learned a ton and it is going to be a fascinating sector to follow in the years ahead thanks so much for sharing your Insight today and and please come back and talk with us again sometime it's been a pleasure I loved your questions James

This transcript was generated automatically from the video's captions and may contain errors.