Constellation Research founder R "Ray" Wang discusses the criteria for the Constellation ShortList #AI and Machine Learning Cloud Platforms, emphasizing the importance of handling large #data, delivering compute power, and supporting contextual decisions.
Ray highlights IBM's Watson X for its clear project onboarding, ethical AI capabilities, and popular use cases like Q&A resources, content creation, and chatbots. Ray explains how Watson X accelerates innovation, reduces project timelines, and supports a maturity model from augmentation to automation.
View the full Constellation ShortList here: https://www.constellationr.com/research/constellation-shortlist-artifici...
---
Full Video Transcript: (Disclaimer: This transcript has not been edited and may contain errors)
Ray, hello everybody. I'm Ray Wong with Constellation Research, and today we're going to talk about the constellation shortlist in artificial intelligence and machine learning cloud platforms, and that list is IBM's Watson x Now why do we have the shortlist? Well, the shortlist covers a number of things. It's aI platforms. It enables organizations and individuals to build intelligent applications based on data. So think of AI platforms as the way to facilitate the ability to ingest complex data, address rapidly generated or constantly evolving data. Rec, define, amplify, hard to find signals, craft models where human powered analytics are slow and able resolution for highly iterative models, and of course, decrease the time to generate models and improve the accuracy rate.
What we're looking for are AI success components, and these include the ability to handle large amounts of complex data, right? So you got machine, you got streaming, you got IoT, you got tons of data coming in, deliver massive compute power, compress the time to get there, provide the right math talent and, of course, embody domain expertise, leverage the human UX and, more importantly, support contextual decisions. The whole point of these AI platforms is to provide the infrastructure to support contextual decisions. These contextual decisions power an array of AI driven smart services, and these are the ability to deliver the next best actions across a whole through whole number of business processes. So how do you get on the list? Well, we looked at 100 solutions. There are about 10 that we really like, and then, of course, we come down and narrow it down to the shortlist, which is even smaller than that. And so what do you have to do? Well, you've got to have cloud based, scalable compute infrastructure. You've got to be able to do compute power acceleration. You have to furnish microservices, application programming interfaces, so the APIs have to be there and really be able to access the algorithm libraries and services, ELT, ETL, all important.
And of course, how do you support multiple llms, these large language models? People want to try different things. They have different use cases for those. And of course, pre tuned AI services are important, and supporting of automation of processes where you can now a lot of this is also the way we train and we actually build inference models. And so how do we scale out and do supervised and unsupervised learning? And of course, there's machine learning, interpretability, monitoring, prediction performance, enabling automatically, automatic visualizations, detecting anomalies, and, of course, creating toolkits for low code algorithm creation where you can and of course providing recommendations, rankings and data labeling, and, of course, support ml op pipelines, very, very important. So this is kind of there. Now, why is IBM on their list? Well, it comes from briefings. It comes from our interactions with customers and partners. We ask people, What are the top solutions? Which ones do you like? And when we talk to people about IBM, what they like as the clear roadmap to onboard projects, it's easy to start building a project. It's easy to manage your data and AI and AI and of course, there's the ethical and responsible AI capabilities where you want to be able to govern your AI models. Now, the popular use cases that customers are using IBM for are building a Q and A resource dump all the information in be able to ask a question. It intelligently figures out the ontologies and tells you exactly what the answers are, and helps you organize information at our pace we've never seen before create content. So drive down the content supply chain, make it a lot easier. So what might take you four months will take you four weeks. What could take you four weeks might take you four days.
And of course, being able to play chat bots creating that interaction that's out there. And of course, we see a lot of work happening in coding efficiently, being able to use any kind of generative AI to do code checks, improve quality, avoid any kind of security vulnerabilities. That's accelerating at a rate, or what might have taken you a week to build a project might take you a few days. So that type of innovation and acceleration is happening. And of course, customers are looking to reinvent their customer experiences. And further down the line, we're seeing the ability to unlock insights. Constellation has a maturity model in the space, and it's really about starting with augmentation, getting to acceleration, getting to level automation, and from there we'll get to agents and advisors. And that's really where we see that five level maturity moving forward. But once again, this is the short list. We're talking about IBM Watson X and they're on the artificial intelligence and machine learning cloud platforms shortlist.
