We just sent you a verification email. Please verify your account to gain access to
Cloud AWS re:Invent Coverage. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For Cloud AWS re:Invent Coverage
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for Cloud AWS re:Invent Coverage.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
Cloud AWS re:Invent Coverage. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to Cloud AWS re:Invent Coverage
Please sign in with LinkedIn to continue to Cloud AWS re:Invent Coverage. Signing in with LinkedIn ensures a professional environment.
John Furrier interviews Dilip, head of Q for Business, about updates on App Studio, a tool for building applications with low or moderate coding skills. The tool aims to enhance productivity and prevent shadow IT within organizations. Q for Business is designed in three phases - indexing data, performing analyses, and enhancing workflows. By saving time and improving productivity, it benefits users in various industries. The tool offers app-building capabilities and connects to platforms like Salesforce. The goal is to simplify AI integration into business wo...Read more
exploreKeep Exploring
What is App Studio and how does it aim to improve productivity for line of businesses by allowing individuals with some coding skills to build websites that adhere to security and privacy practices?add
What are some examples of companies using Q Business for different use cases and how is it helping them save time and improve efficiency?add
What are the different phases of search and how can it lead to increased productivity within a team or department?add
What are the popular areas for app creation and distribution right now?add
>> Hello, welcome to theCUBE's coverage here. I'm John Furrier, host of theCUBE here in Seattle at the headquarters of AWS. It's the re:Invent building. Dilip is here. He's the head of Q for Business and other things. Good to see you again. Thanks for coming on theCUBE here-
Dilip Kumar
>> Good morning.... >> in your home turf.
Dilip Kumar
>> Yeah, thank you for having me.>> So Q for Business, you've got some news, pre-re:Invent. What's the news?
Dilip Kumar
>> Yeah, so one of the things that we previewed in July at the New York Summit was App Studio. We have low code, we have pro code. There's this middle category of low code tools and generally what tends to happen is that line of businesses have a lot of particular needs for very peculiar needs for applications that they need to get built. But it turns out that centralized IT and engineering is usually backlogged on a whole lot of things that they need to do. And so most of the things that line of businesses need tend to end up falling below the line, which impacts productivity. So App Studio was introduced as a way for how do you allow people who have some amount of coding skills but not a lot to be able to get reasonably productive and not just improve productivity for themselves, but also to be able to do it for their own team or their organization. So using very simple generative AI prompts to be able to build websites that has a reasonable amount of punch, but it also adheres to all the security and the privacy practices that IT administrators are looking for so that we don't have shadow IT perpetuating toward the organization.>> In New York, that was a moment where you saw the light bulb go off in everyone's head and they say, "Wow." And I was just talking to Swami and I talked to Matt Garment about it a little bit, but Swami really hit that point home was the productivity gain now in the business lines is where you're seeing the step function in value. And the labor and the cloud initially was developers, right?
Dilip Kumar
>> Exactly.>> Okay. Developers got that too with CUBE. CUBE is very popular with another CUBE developer And now on the business side you're seeing the productivity of that labor. So the TenX engineer was cloud. That still continues to be the case and now we're seeing productivity like a TenX business person. So the process and the domain expertise in the business side is being impacted by Q for Business because the workflows impact their job. So you don't want to write SQL code? Well, maybe the AI could help. So this is where the workflows come in. So can you explain... It's super exciting area and by the way, agents are coming too, right there too. So take us through where you are on the progress bar of Q for Business to impact the worker who's sitting there saying, "Well, I can actually do a little bit of the heavy lifting myself and I don't have to get engineers involved. I'll write the query or I'll do this." So where is that going?
Dilip Kumar
>> So when you think about Q for Business, we think about it in three phases. One is organizations have this treasure trove of data. They're all usually in silos. Different people have access to different things. You mentioned SQL. SQL used to either have to write SQL queries against databases if you knew it or you had to go to a BI engineer to be able to get the level of productivity you needed and that was then available in BI tools. What's happening is that you don't come into work on Monday and say, "Well, today is the day I want to work on getting information from databases and tomorrow is the day I want to get information from my documents or SharePoint, or Confluence, or Wiki."
What you really want is that you have problems and you want access to the data wherever they are. As long as you have access to it, Q should be able to provide access to it. So the way that we built Q is to say why not use all of the information that companies have indexed. And in many different places, bring that all there but also cover both structured as well as unstructured data and be very permissions aware. I mean, organizations have spend a lot of time and energy trying to be able to figure out as to how to be ACL aware, permissions aware for people who should have access to the data, should have access, people who shouldn't. And so when we index all of this data, we get all of it, make sure it's appropriately permissioned so that the right people have access to it. So if you didn't have access to any of this information outside of Q, you won't have access to it in Q as well. So that's the first part. The second part is that what you really want to be able to do is that not just post questions and get information, you also want to be able to do analyses. You also want to be able to take action. So converting this from a system of information to a system of action is super important. And then the third aspect is that we always talk about this notion of where do people work and what's the kind of work to do? You want your assistance to be available where people are doing work. So if you're working in Slack, it should be available there. If you're working in your line of business application, it should be available there. That's the general idea. So if you do those three things, it turns out that people know that they can rely on assistant. People know that they have access to the data that they need and that they have a bevy of actions that they can take to do their day-to-day jobs. That's really helpful.>> Yeah, I think the product too is great. Talk about some of the data you guys are seeing? What I love about Q is Q for Developers is one, it helps the coders but also Q for Business is serving some use cases that you guys have done. Talk about some of the improvements and data points you're getting out out of the workflow.
Dilip Kumar
>> So I'll start with the developer use case and then move there. Even developers need access to a lot of documentation. They have tons of documentation. This documentation is usually buried in different places. We're seeing that people are posing millions and millions of questions and saving 4,000 hours of work just not searching for documents. That's just combining Q Developer and Q Business is providing that benefit. I'll give you some other examples of the kinds of things, some which you would've normally expected and some atypical even for us. So most people say that customer support or support use cases are very, very clear places where you can draw on a vast amount of information, index it so that when queries come in you users have it. So we have companies that are using it for onboarding. Smartsheet has the Slack application where you can do at Ask Me. And what tends to happen is that new hires come in, they pose a whole bunch of questions, typically the community answers, but now you have Q Business behind the scenes which is indexing all the company docs and being able to provide answers. So they're getting answers in a fraction of the time that it used to take. There's companies using it for recruiting and saying that when resumes come in, why don't I pull out the relevant content if this person passes the types of job descriptions that I'm responsible for? Why not send it to the right hiring managers? There's a bunch of that automation that is happening, which used to be someone's daily job to come in and have to do a lot of grunt work. The interesting examples would be things like a manufacturing company used to send out support technicians to factory floors to be able to take care of their equipment. The problem is that when you have a huge churn in your own support staff, most of the people going out to service these machinery are not very well-equipped with your company's manuals and the machinery that's out there. But now that they can have Q Business at their fingertips, it allows them to be able to guide them to be able to fix this. So the difference between an expert technician and a new one->> We've all been there gone....
Dilip Kumar
>> has gone north.>> How do I fix my sink? I go to YouTube. Okay, there it is.
Dilip Kumar
>> Exactly right. I mean, we just had an electrical problem in our house and I was like, "How do I change this? How do I do that?" It's kind of interesting.>> You bring up a great point because I don't want to categorically just say enterprise search, but if you look at data, people need data to do their jobs. Data is in application. So search concept is defined what you're looking for when you need it contextually and matching it to the right context. So that's just search. So search is a killer app here because the value proposition is whether I'm searching for an answer if I'm a developer or, "Hey, what's the quick... Where's that documentation to, hey, I want something out of the SQL database. Write me some SQL code." I mean, this is just the user experience with data. I mean, this is what we're talking about here.
Dilip Kumar
>> Exactly.>> Search tends to be the low hanging fruit. But search is first and then integrating that data through discovery. I mean, I guess discovery, I mean this is how it is, right? I mean how do you frame this into a simple concept?
Dilip Kumar
>> Well, even search, if you think about it, you're absolutely right. Search comes in multiple phases. One is I know exactly what I wanted. I know when I find the answer. I get the right answer. It's an authoritative solution that comes in that's been indexed. But then there's other scenarios where the search is just the starting point. You start with the search. You say that, "Well, I want to be able to figure out what my weekly sales are," or to be able to create a report for weekly sales and it produces a weekly sales. But it turns out that every Monday you have to go to your boss or you have to go to your team and to be able to indicate what your sales operations are doing. Why not automate that and why not build? So we had this concept in Q Business known as a Q app which allows you to take conversations or things that you would normally do on a much more repetitive basis, automate it so that the next day when you come in or the next week when you come in on Monday morning, you come in, you don't have to sort of start that process of typing it in. You just open the app. The app is updated with the latest information. You can take that same app now and share it. So I can share it with you. I can share it with somebody else and my team and now it runs for their customers and their app. So now we've gone from something that I search and I get information to. I then using that to be able to create applications and that I can then automate and then distribute to a team that they can use. And so you go very quickly from individual productivity to team productivity, to department productivity. And I think that's where the power truly comes in.>> What's interesting is gen AI brings us to a whole other level of intelligence to keep things like closing the discovery loop, which you're talking about there. But also knowing things like concepts like memory. Not like memory on a chip, but where was I in my work? I'm going to go home now or I might be in a different form effect on my mobile device. So you have all this connected tissue with data. This is the value that we are going down that road. So as a worker, it's not just discover the answer. I need the answer now to go to the next step.
Dilip Kumar
>> Exactly.>> And now I'm in a progression. I'm in the workflow.
Dilip Kumar
>> Like an assistant or a teammates. It's exactly the concept of memory. They know where they left off. They can pick up from there. They have contextual history as to the kinds of things that you have done. Over a period of time you can see very normally that these things are going to be suggesting and doing work on your behalf rather than you having to initiate it.>> So how do I set this up? And I just think about myself as a use case. So imagine myself, "Okay, I'm sold on Q for Developers." Swami sold me on that. Well, Deepak did a year ago. I see the value there, development side. On the business side, I could see multitude of benefits.
Dilip Kumar
>> Yes.>> How do I get started? Do I need to set up a bunch of data? Do I usually use the App Studio? Take me through if I'm a customer onboarding it.
Dilip Kumar
>> Yeah. It's actually super simple. It's four steps. Four very easy steps. You can just onboard your data. You can point to what is your unstructured data. You can point to your wikis, you can point to Confluence, you can point to different types of company data that exists. And very simply the admin can just create an application and then they can invite users who need to be onboarded to that application. What tends to happen is that one of the benefits that you see is that you can start getting value from an->> A business person can set this up?
Dilip Kumar
>> Like an administrator. An administrator and a company can set it up. It's a very simple process. All it requires you to know as to what is the content that you're trying to index that you're bringing in. What's the user ID password, the credentials necessary for it? And then once the index and the data comes->> It could be S3 buckets.
Dilip Kumar
>> It could be S3 buckets. It's like we have 40 plus connectors to almost all the popular things that we can think about.>> Basically the location of the data.
Dilip Kumar
>> The location of the data and some credentials to be able to pull that information inside which few people have. Once that comes in and it's authorized for your users, your users can start getting utility on day one. And the reason for that is that they don't even have to wait for all the data to be indexed because people have content on their laptops. They have files. They have other things that they get done. You can start chatting with Q Business to be able to get and summarize content, generate content information right away. And as your company information starts coming in, it just augments it. So the things that you can end up doing gradually increase over time. You can create Q apps with your company data. You can start taking actions on your company data. You can upload, summarize, and deal with anything that you have. That's the benefit of it. So there's the admin-related work to index all of your content, but you don't have to wait for all of that content to be ingested. You can start getting utility right away.>> Yeah. I mean, I think search is such a low-hanging fruit on this use case. You can start just index everything.
Dilip Kumar
>> Exactly.>> And then the workflow end-to-end seems to be where the agents are going, the agentic systems. I mean, the hype cycle on agentics is really high right now and reality has come downstream when see... Obviously, because you've got to get the infrastructure and data layers all set up. What does that pre-agentic wave look like? Unstructured data, semi-structured data, structured data, all good to go or...
Dilip Kumar
>> When people talk about agentic information, they're usually talking about things that are autonomous that can take action on your behalf. We're already headed in that direction. So you can start with simple actions that are pretty deterministic. I just mentioned Q apps. Q apps is a way where you can automate things where something is taking action on your behalf and then over a period of time what the continuum that we are seeing is that the complexity of the queries that people are posing to an assistant is increasing and the ability for the assistant to be able to decompose that into very specific problems and then have multiple different applications or sub-applications or agents as you will handle that and then compose the answer back to the user, that's definitely headed in that direction. I think the key is the orchestration and figuring out how to compose the problem properly.>> Talk more about the Q apps. I think that's fascinating. So take us through where you are in that. Are people building apps today? What are the use cases you see? What are customers telling you and how do I build a Q app?
Dilip Kumar
>> Yeah, it's actually... It's trivial. So once you have Q Business, let's say you have a conversation, you can just type in and say that we talked about like, "Oh, I have weekly sales," and you could just say, "Create a Q app for my weekly sales and do this," and it just guides you through the process. It's a very simple process. There's no code involved and then you can select different things that you want. You can select who you want to send it to, what's the frequency at which you want this to run? There's a few knobs that you have you could select. There's eventually things like model choice and a variety of other things that you want certain things for image generation or certain things for text. But it's effectively what tends to happen is that in two or three steps you can create a pretty powerful app and then this app can then get distributed to anybody that you want to publish this to. And then you have a library of apps that are constantly available so that anytime that you come in to your Q Business console or the main place where you're interacting your assistant, you have access to all the apps and not just you've built but other people have shared with you as well.>> What's the popular areas right now? Is it sales information? Is it some of the basic stuff? What are some of the-
Dilip Kumar
>> It's actually across the board. It's super interesting and this is where human creativity is. Salespeople use it for a lot of sales-related tasks. We have HR apps that are being created. We have apps. One of the most popular apps at Amazon is this app called . We do this whenever we have it started off with engineers doing correction of errors for when they have issues that happen where you say, "Why does this happen? Why does this happen? Why does this happen?" It turns out that people are actually using it for problem-solving and designing things. So when they want to interpret a new design or create alternative designs, they create multiple different versions of it. You have apps being created for resume searches. You have apps being created for interview questions. You have apps being created for operations-related things. People have created apps for private pricing, agreements that they have.>> I saw someone create an app for answering the interview questions. You even don't think about that, and then they read the answer, they get hired.
Dilip Kumar
>> That's kind of interesting.>> It's kind of a funny meme going around on that, but that brings the point that now the data is going to accelerate things, answers and discovery.
Dilip Kumar
>> Exactly.>> I want to ask you about Salesforce and other integrations. You mentioned sales. Is Salesforce partnering with you on this or what's sales? Because connecting to sales data
Dilip Kumar
>> We have connections in Salesforce. One of our internal applications that we've sort of built the sales team built, it's called Field Advisor. Field Advisor uses Salesforce data and it uses Q Business together in order to be able to answer the queries that you want, Salesforce. And this is again a perfect example of salespeople have access to certain amount of information in Salesforce, but you also want to be able to do something with that information. You want to be able to create an email for your account for an upcoming account that you're preparing or meeting, customer meeting you're preparing for. Or you want some information about some historical trends on your customers.>> So you can connect today with Salesforce?
Dilip Kumar
>> Yeah. We can sort of connect to with Salesforce. You can ingest that data. Q Business can front several elements of it.>> This is the key to this connected ecosystem that's development. I'm calling it not just API connectors, but actually data going back and forth between systems.
Dilip Kumar
>> Exactly. That is exactly what people want. What they want to be able to do is to say that I want access to the data that I have. I want to be able to take actions and those actions can go and write information and systems that I'm connected to. So you have connectors and you have actions. And from a user's perspective, they shouldn't have to deal with the muck of having to figure out all of that. There should be something that sort of abstracts that away which is what Q Business does.>> What's your partner's strategy to roll this out? Because Q for Business is clearly going to be the layer for interfacing to the business, making it essentially businesses code. I mean infrastructure's code was great, but that brought DevOps. Now you have businesses code.
Dilip Kumar
>> One of the things that we've seen in almost everything that we have done, when I said that there were three things, the idea was that data has gravity. So it's very important that we get to the data. It turns out that a lot of companies have the data in AWS, so it's very helpful. We also have the largest ISVs also in AWS. And so our strategy with Q Business has always been that we want to be where people are doing their work. So if you're in an ISV application, the presence of Q Business indexing the data should help your presence in that ISV application. If you are using a line of business application, it should be applicable there. So our approach has always been to be able to figure out as to how is it that you can get utility to our customers no matter where they are working, but also indexing all the data that they find that is relevant to them with the right permissions. So it actually works in both ways because ISVs get the benefit out of it. The company gets the benefit out of it and employees or their customers win.>> I mean, I was talking with Swami and joking with all this CUBE transcripts on S3, every single CUBE interview since 2012, just point Q Developer at it, Q Business and I could have CUBE apps.
Dilip Kumar
>> Yeah.>> I mean technically possible early on S3.
Dilip Kumar
>> That's a hundred percent possible. It's totally doable.>> Talk about the magnitude of this inflection point. I mean, you've seen pre-gen AI. You know how hard it is to do these workflows. I mean the old days, five years ago, 10 years ago, you'd have to have a big project plan if you want to integrate and create a system, an application on top of these enterprise workloads. It's complicated. The enterprise has all these knobs and buttons to pull. It's not just throw AI at the problem. You can't just throw AI at the enterprise. You got to have what's the identity of this database? There's all these buttons and knobs that people know about, but you can't just have a machine learn it. And to now the old way and new way scope the order of magnitude, step function changed.
Dilip Kumar
>> I think it's monumentally different from the old way versus the new way on several dimensions. One, I think administrators are sort of realizing that they have a lot of control on the way that they want to be able to construct and ingest this data. It's making it a lot easier for them. We're also introducing an entirely new class of people. There's making people efficient in the way they work and then there is finding new ways of doing things. So what's happening is that you're taking 80% of the things that people used to do and you're saying that we can actually do it in a much more efficient and a much shorter period of time, and that 20% of things that you used to do that is really valuable, you can make that 10X better. We're actually seeing productivity gains happening as a result of both of those things happening, which I think is reasonably unprecedented. When you sort of look at several of the shifts in computing that have happened over a period of time, which is what is interesting. And a lot of this was originally text-based. It's going to other modalities. The complexity of the types of things that people are asking is increasing. So it's a golden age.>> Productivity is the killer app. One observation I'll share with you, and I've talked about this in the CUBE pod a little bit, but the word harmonization is interesting because you're harmonizing data across different data sets and understand that with gen AI. But one little behavioral change I've seen in companies that have done this right with Q Business have this mindset is that this more... Morale is better because the developers and the business people, the conflict goes away when someone can just write their own SQL query by saying, "Hey, I want access to this data." The person doesn't have to do that rock-fetch mundane work.
Dilip Kumar
>> Exactly.>> Oh, they want another request from the business person. Or I'm not almost simplifying it, but the relationship people are collaborating differently. The psychology of the interactions are different. So you're seeing more harmony in the roles of people.
Dilip Kumar
>> I like that phrase. It's kind of interesting. They were always on the same team but pitted in some shape or form but now they're sort of rolling the same->> It's friction. And friction goes away, and that makes people happier.
Dilip Kumar
>> Totally.>> I mean that's a productivity benefit. It's funny, I was talking to Swami about the benefits of Q in developer, and one developer told me with gen AI. He's like, "I have more beer time," since he wants to hang out with his friends. What he means is I don't have to spend all that time doing things.
Dilip Kumar
>> Yeah, I think it's exactly what I said. It's when you remove a lot of that undifferentiated heavy lifting and a lot of the undifferentiated heavy lifting on the business side wasn't collecting the data, and getting the data or having access to the data. Less time was spent. They would trade spending more time on more thoughtful analysis any day of the week if they could. And this is allowing them to do that. And it's allowing people to have agency and a little more control over how they spend their time, which I think is incredibly liberating and harmonizing.>> I was talking on the CUBE pod with Dave Vellante on this and I said, "The business people are the new IT," in quotes. And he said, "What do you mean?" I go, "Well, IT used to be serving the business. They would provision the desktop to your cubicle and get you access to all the databases. They had to bring technology to serve the business. And then technology became the business. Now, the business people that know the workflows are serving the business through their knowledge. And so they become the new IT value because they're using technology better. They're not just consuming technology, they're using it.
Dilip Kumar
>> Exactly.>> And they're productive with it.
Dilip Kumar
>> Right. And I think it's the notion of English becoming the coding language of the future. If you can describe the kinds of things that you want, the ability to be able to do those kinds of things is super >> Personalization, having memory in your workflow.
Dilip Kumar
>> Exactly.>> It's mind -blowing.
Dilip Kumar
>> It is. It's a good time to be alive.>> It's great to be here in the Amazon Studios. Dilip, thanks for-
Dilip Kumar
>> Thank you so much for having me... >> coming on. Okay. This is theCUBE coverage here in Seattle at the headquarters of AWS in the building that's actually called re:Invent. I'm John Furrier, the host of theCUBE. Thanks for watching.