Thoughtworks Haiven goes past coding by integrating AI into program advancement lifecycle

00:00 [This transcript was auto-generated.]
Hello, all people, welcome to DEMO, the demonstrate exactly where we have organizations come in that present us their newest products and features. Currently, I am joined by Mike Mason, he is the chief AI officer at Thoughtworks, repeated visitor on our other displays. So welcome to the DEMO exhibit. Many thanks, Keith. Fantastic to be below. And you are gonna demonstrate us the Haven crew assistant, suitable the and notify me what is Haven and what does it do.
So generally, there’s a ton of hoopla suitable now in generative AI close to what you can do with it when developing software, ideal, specially code generation. So these AI’s are spitting out tons of code for builders. But we essentially believe that generative AI is valuable throughout the software lifecycle not just for coding, for all the other points that the software program workforce does when they are setting up, creating program. And Haven is a crew assistant to aid you throughout all of people unique roles and duties. And I’m likely to present some of these today.
Who is the primary form of purpose in just a firm that would that would reward most from working with Haven? Is it across the business? Is it just builders? Like who can this support?
So it really is software teams, so it really is everybody who will work on, you know, setting up software program to develop some enterprise aim. So, you know, that could be from people who are carrying out the needs for the software program, people who are building consumer interfaces, people today who are in fact undertaking the coding, testing deployment, functions in output. So like the the total software package lifecycle from form of plan to
production? Yeah. Alright. And so let us soar in, then you will find quite a few characteristics this has, I assume you confirmed me a slide the place it had bullet record right after bullet lists just about every bullet list, but we’re gonna aim on a few awesome features from from this. So let us go right into the demo.
Positive point. Okay. So the initially factor to communicate about is the simple fact that Haven is definitely about assisting groups do some thing we get in touch with understanding engineering. And so what they are performing is figuring out how to specific to an AI, what is the application that we’re building? Why are we setting up it, what is the technological architecture, all of that kind of stuff. And so you can find a feature in in this article referred to as Group information. And 1 point I’ll say about this is that you can find a lot of demo the place listed here, the intention is for persons to consider this software and customise it for their have teams. So you happen to be gonna see stuff running in opposition to GPT 3.5, GPT 4, this thing also works with Gemini on on Google and also on the Claude on the AWS.
Is that truly important that a business has accessibility to multiple MLMs?
They really don’t have to have obtain to a number of, but it assists that this can operate with whichever a person they experience relaxed with. Since, you know, with AI, there’s all these mental property fears, all that variety of stuff. So if a business has now figured out, Hey, we are relaxed making use of this individual procedure, Haven can work with. So within listed here, you can find this demo that we’re going to be utilizing currently, which is about EV charging stations. So electrical vehicle charging is like the challenge domain that we’re heading to be exhibiting off. And in this, you can say, hey, you might be part of a crew constructing software package for EV charging stations. And this is variety of the varieties of parts. This is why we’re making it. These are the functions of the computer software, all of that form of things. So generally, you’re you are teaching the AI, what it is that you happen to be building, so that it can interact with you in intelligent strategies. The other issue we have acquired in in this article is Staff file centered understanding. So OCPP is a variety of specialized documentation for EV charging. And so this makes it possible for you to obtain that as effectively. Okay. So the initially issue we are going to do listed here is demands breakdown. So this is in essence getting a significant amount plan of some performance that you want to construct in your software package, and then breaking that down into needs that a developer could really go and carry out. And we’re likely to use an AI to help us with that. So the tale we’re functioning on right here is applying a secure and adaptable payment processing process for EV charging. And what the AI is accomplishing listed here is it is really having that description, additionally all of the other expertise all of that context that we have set in, and he’s coming up with ideas for capabilities in the computer software that will be valuable for, for this distinct story. And so it truly is it truly is starting off to produce these points, Hey, possibly I need to be in a position to sign up as a user, I can use unique payment approaches, all of that type of things. Probably you want to see real time charging status. That’s a acceptable notion. What’s happening in this article, ideal is the AI is aiding anyone feel about that demands. So not just a thing which is apparent, but like coming at it from several distinctive instructions and having that sort of AI creativity to assistance you out, alright. And then what you can do is you can you can basically check out these factors in a little bit extra element. So you can say, Please give me much more aspects. And the AI will truly like go to city and flesh that out into something that is finding a bit bit a lot more full and finding one thing nearer to what you can employ. It’s also asking important issues like what types of payment mechanisms do you want to just take in this in this individual scenario? You can even then I’d say, just remember to make your have assumptions.
You will not usually normally want to see that with it with AI.
Properly, I imply, I consider you, it’s appealing, right? Like, you probably would not just do this if you were being doing this for serious. But for the needs of the demo, we can get the AI to fill in all the blanks for us, and occur up with that that functionality. All proper.
That is great. And then this the 2nd a person, you know, I found you have a diagram in this article. I assume this is a interesting aspect of the diagram evaluation or diagram summary. Yeah. So in essence,
we’ve obtained a diagram below, this is in fact from AWS, just a random diagram of how an EV charging station might do the job. And these are all type of the complex elements to do with that.
And if you handed that to me, I would be like, I have no notion what that is. So you’re
not specialized, no man or woman, so you would not know it wouldn’t know what that is so. So we can do anything named technological diagram, description. And we can upload an graphic, right, which is that very same one that I was exhibiting you just now. And we can describe that and we can say, you know, this is an AWS architecture for EV charging. And what it really is carrying out right now is it really is going through the this is using GPT for vision. And it truly is form of on the lookout at that graphic, attempting to evaluate it realize what is truly in that. And once it really is finished that, we can then get the system to reveal it to us. If we are non specialized, and we do not, we don’t know just what that what that point is,
I am gonna go off script just a tiny bit. But if you uploaded a image of a cat, would it just be capable to explain? It would probably say inform me about the, you know, technological architecture of the cat? I am not guaranteed it would give you nearly anything specifically valuable. Alright, there we go. Alright, so did that. So it truly is generally described the complete, you know, as it can be conversing about the diagram. And then what I am performing in this article is I am indicating, hey, make clear this diagram to me. And it states, Absolutely sure, excellent. And it tells me about the EV charging level, a load balancer, vehicle scaling all of these AWS factors that had been on the on the diagram there. And fundamentally, which is, you know, you can then drill down and request for additional detail. You can also do issues like saying, hey, I want to put into practice this unique element that the payment processing. And you could say to the API, which architecture elements are applicable, proper for that, and it will help you realize that a little bit superior? Yeah,
as we’re going through some of these attributes, what would a user or a member of the computer software workforce do in advance of a software like this? What would they have to do? Would they just be accomplishing a good deal of other research? Would they have to count on other teammates to explain issues? Yeah,
which is a fantastic dilemma. They would normally have to count on like the most pro person on the staff to recognize what are all the things on this diagram? And what do they necessarily mean, the wonderful detail about building some of that understanding into the AI is that you can then have some of that benefit of conversing with an professional, without requiring your group qualified prospects to expend all that time with you. And you can use this for all the diverse roles on the team. So somebody who’s an analyst, like trying to determine out what characteristics are practical? They can they can do that. They can profit from that precise very same form of specialist AI.
Proper. Okay. And then you can also then, just after you’ve gotten the answer, you can follow up with thoughts, but I want to get to the future the future 1. I think it truly is referred to as Knowledge chat?
Yeah, so mainly, so I was speaking about the significant document with all of the technical specs for this specific protocol. So you will find a bunch of distinctive OCPP Yeah, accurately suitable. But you can develop other things in below. So for instance, we have bought you know, you can glance at the Thoughtworks guideline to agile software package shipping and delivery, and use that as your know-how foundation, and then ask inquiries about that if you required to. We’ve also obtained Martin Fowler’s internet site in listed here. He is our chief scientist. But if I seem at the OCPP protocol, right, and I’m just heading to deliver that up for you, briefly right here. So this is a 580 web page, super dense, specialized document, appropriate? No way anybody’s ever gonna go through 580 web pages really worth of things. It’s bought like architecture, diagrams, all sorts of stuff in that, ideal. So the place of this is to be in a position to ask a query. So in this article, if I’m doing the job on that payment processing tale, and I’ve acquired this big architecture doc, I want to realize, is that really practical. So given that story about payment processing, is this OCPP point likely to support me at all. And so it is saying, not genuinely acquired any specific data. But there is anything about credit rating cards and debit cards in this darkish. And if I scroll down a bit, and I have a glance at some of these points, you can see it truly is got webpage numbers, it is heading to drill me into site 130. And if I zoom in a minor little bit on this, you can see that this is an case in point of one thing in this technological doc that has to do with payment processing.
So it would scan the entire document and then based mostly on the issue that you request, it can notify you specifically where by
it can notify you is that just a search term lookup is not key word search? No, it’s basically generative research. So it is really accomplishing a valuable form of, you know, genAI design accessibility to this expertise. You know, we have got many different knowledge bases loaded into right here. 1 of them’s for a health-related document process, which has like, you know, 2000 internet pages worthy of of their their operational documentation for how to run the factor. And you can query that thoughts as effectively. Yeah.
So we got just one far more function that you want to display this is about menace modeling or brainstorming.
Yeah. So this is type of modeling. And I’ll kick it off. So we can we can speak about it whilst it really is operating. But basically, I’m chatting about who’s using a element, what are the assets that we are hoping to guard, and around what’s the info stream going on in this article. And Danger Modeling is essential, for the reason that you you should not want to do safety just at the stop of a challenge. Preferably, you would like everybody on your workforce to be pondering about stability, particularly developers, but it really is form of a specialized exercise. And so what we can do in this article is we can get the AI to aid a developer, think about risk modeling for the characteristic that they are implementing. Okay. And this is beneath the addresses, this is making use of a distinct design and style of risk modeling, identified as stride wherever you assume about certain forms of threats to the system. And so listed here, the AI is coming up with matters like spoofed identification with a stolen credit score card, tampering with payment, details, all of that type of things. So these are potential methods that there could be some variety of a menace to the procedure.
And this is particularly about the EV charging just, since
we have offered it all of that context about what the crew is developing, also the certain element that we are performing on, and then the AI is encouraging us look at that unique threat. And then you can go into investigate these matters. And you can say, could you give me extra particulars and mitigation? So this will really form of demonstrate what is this danger? And how could possibly you make ways to perform close to it?
That is definitely awesome. Ok, and so, is there a pricing model? Is this subscription based mostly? Okay, so or is this Do they implement certain areas of the system.
So this is at the moment, this is an accelerator for ThoughtWorks customers. So if you’re a client, we can deliver it to you and you can use it, we’re going to be releasing an open up source version of it. And we’re essentially likely to be splitting up the computer software from the understanding pack. So the expertise pack is all of that embodiment of, you know, how you, you want the AI to prompt you. And we are essentially heading to be embodying sort of 30 several years value of ThoughtWorks abilities in how to build company software program into a awareness pack that is going to be proprietary, that will not likely be free of charge. But the primary application is likely to be free of charge.
And just given that you mentioned it, can you reveal to another person who may well not know what Thoughtworks is? Positive, yeah.
Pleased to you know, you do not just create computer software for the heck of it. No, we will not just produce, it can be like no, and I like to say we are the greatest tech corporation nobody’s ever heard of, because we are 10,000 folks globally, we establish a ton of software program for our consumers, you know, airlines, ecommerce startups, or, you know, the comprehensive gamut of organizations. So for men and women that want more details about Haven, if they are Thoughtworks consumer, where by can they go for more facts, you can go to And we’ll depart some links, probably in the movie description. All proper, Mike Mason. Thanks. Many thanks yet again, and many thanks for the demo.
Content to be listed here. Many thanks.