watsonx Orchestrate on AWS value proposition | Amazon Web Services

Key Takeaways

Hello everyone, my name is Eduardo. I'm a partner solutions architect with AWS. Today with me is Karan. Hello Karan. >> Hi Adu. Hello everyone. My name is Karan. I am part of IBM. I manage IBM AWS partnership from a data platform perspective. >> So Karan, you know nowadays with all this generative AI bus, there's a lot of things going on happening so quickly. Why should customers in invest into building generative AI workloads into building agentic workloads? What benefits do they get from this and what uh opportunities are out there? You know, I would like to start with the scoop. Whole of knowledge work market is around 15 to 20 trillion dollars in globally. >> Okay. >> Okay. That's the knowledge worker market size. And even if we are able to save 15 to 20% of this market size by using AI agents, that's a tremendous value to our customers. >> Okay, >> it it turns out to be millions of dollars. So if say I take 15 to 20% of saving, that's around dollar 5 trillion market for us. >> So we are seeing uh customers deploying AI agents to make it more efficient throughout their organization. >> Okay. Now how can these customers then start building their AI agents you know to go towards this cost savings and optimizing their organizations on AWS. What type of solution does IBM has to our customers so that they can start in this journey. IBM is a family of AI enabled tools like Watson X platform. Watsonx orchestrate is key part of this Watsonx family. uh What's next orchestrate is available on AWS as a fully SAS offering managed by IBM and customers can start using that their own their organization ASAP. >> So without any deployment because it's a SAS solution they just have to go to the AWS marketplace subscribe and start using it. >> Absolutely. This helps customers build their own AI agents, govern them and then manage them at scale. >> Okay. So what are some of the u challenges that customers can overcome? You know, what are the the u the business results also that they can get by using the Watson X orchestrate capabilities to build AI agents. >> One of the biggest challenge I have seen with customers is that they start a pilot, they start an experiment and they're not able to scale that to the production. So I would say pilots they will start hundreds of pilots or they will start a one strategic pilot >> and how do they take that into production in a in a scalable way >> okay >> that's the biggest challenge I see and with what's next orchestrate with no code platform customers are able to do that scaling part seamlessly >> okay so customer even business users is because you mentioned no code they can start developing agents but if who if you also have a persona like more like a software developer or an AI engineer, they also have the agent development kit on what's an X orchestrate. And I think one of the cool things that IBM has done is you have a model gateway that allows customers to choose whatever models they want to use in the back end. They can obviously use the IBM granite models that come out of the box, but if they want to use things like cloud running on Amazon Bedrock or even Granite deployed through the Bedrock marketplace, they have the ability to do that using the ADK. I I would say Edu LLM is the heart of the whole agentic AI thing and with gateway you have talking spoken about customers can use grey knight on bedrock as a heart of this AI agents which they are building or they can use any third party AI models as well like cloud A or they can use other models available in market. >> Now on top of what we've discussed here what are some of the other benefits the customers get? So you see why they are not able to scale pilots from production. One of the biggest inhibitor is skills. >> Okay. >> So business users have different set of skills. The IT users have different set of skills. >> So with with no clear platform, business users can also use it with the gateway you mentioned. The developer can also use skills to you know develop a agent. Second thing enterprises have hundreds and thousands of applications. How do you integrate with them seamlessly? >> Okay. with uh inbuilt integrations by using MCP and A2A protocols. You can integrate what's next orchestrate with different applications. Okay. >> Like uh SAP success factors, work workday and and so on. Right. >> And you can create agent to agent communication using these protocols as well. >> Absolutely. Absolutely. And and third is scale and when when you have that those protocols in there, you can scale AI agent throughout your organization in a cost-ffective way. So Karan, can you tell us about the business outcomes that customers can expect by using what's on X orchestrate and embarking on this AI agent journey by deploying AI agents? Now they have more time to innovate. So innovate faster. That's one of the biggest business outcome I have seen our customers realizing. I would say that you know innovate faster. One of the biggest business outcome I have seen. The second outcome I've seen is around accelerated time to value. Okay, rather than waiting for months to get value or years or 6 months to get value, customers can realize value in weeks time or days time. Okay. >> By building AI agents inbuilt in what's next software, >> you can use all the tools, all the no code capabilities to develop agents quite quickly. >> Absolutely. And third is around costs. Now with all the mundane task automated customers can save lot of costs. That's the third outcome I've seen with customers. >> So you you can automate all the boring tasks, repetitive tasks and you start focusing more on innovation, creating new products, new solutions, increase customer experience, right? >> Yes. >> Okay. And what are the top use cases that you see customers applying AI agents into? I have seen many of my customers deploying AI agents in HR space. That's one of the topmost use case. >> Okay. >> Okay. Uh like doing performance management, doing customer employee onboarding end to end by deploying AI agents. They can take care of this end to end. Okay. >> I've seen that use case really happen in globally in all all our customer space. The second use case I have seen is around customer support and customer support not only one physical channel customer support throughout different channels. >> Okay. >> Uh even even on the call center space as well. >> Okay. >> The third use case is around procurement. >> Procurement it's it's a very pecular use case. Customers can take care of the even RFP release by using AI agents. Customers can, you know, evaluate those RFP responses by using AI agents. Customers can also generate invoices, >> okay, and and pay invoices using AI agents. Now, with what's on next orchestrate, you have all this ability to build your own agents and do agent to agent communication and orchestrate workflows, but you also get these agents out of the box, right? To handle these specific domain use cases. Is that correct? >> In fact, I would say that these three use cases are out of box. out of box >> within Watson orchestrate you'll get HR AI agent you'll get customer support agent you'll get procurement agent >> okay >> if you're going beyond this there is a very easy interface agent builder within what's next orchestrate which you can use to build your own agents and deploy in production >> and because it's running on AWS also as part of your your capabilities you have integration with different AWS services >> absolutely different as services so customers who are using AWS natively they can use orchestrate as well natively >> okay and So, how can our customers get started with Watson X orchestrate on AWS? What are some of the next steps that they can take? >> Super simple. Watson X orchestrate is available on AWS marketplace. >> Okay. >> Customers can subscribe from there and start using what's next orchestrate immediately. >> Straight away immediately and because it's fully managed, they don't even have to deploy anything on their AWS accounts. Right. >> Absolutely. I think that super simple it gets. >> All right. So, this was very informational. I think there's a lot of value. I really appreciate this conversation with you Karan. Thank you very much for being here today. >> Thank you for having me. Appreciate that. >> And see you next time. Thank you. Thank you.