Alexa Carlin (00:27): Hello and welcome to Accelerating Your AI journey. I'm your host, Alexa Carlin, and today we're talking about unlocking the power of AI with the latest solutions from independent software vendors in Lenovo AI PCs powered by Intel processors. Joining me today is Cedric Courteix, Senior Business Development and ISV Engagement Manager, Commercial AI PC at Lenovo. Welcome.
Cedric Courteix (00:56): Thank you Alexa.
Alexa Carlin (00:57): Yeah, thanks for being here. So just tell us a little bit about you and what inspires you to work with emerging technologies.
Cedric Courteix (01:05): Yeah, so I found myself with this role at the cross way of collaborating with very innovative partner on a topic that is moving at the speed of light. It's amazing to see, now that I've been a year at Lenovo, how many partners and scenarios we've been able to explore. And I think it's all based on the joint technology that we are building for our customer.
Alexa Carlin (01:36): Yeah, definitely exciting times.
Cedric Courteix (01:37): Yeah, it is. It is.
Alexa Carlin (01:39): So there's been a lot of discussion that AI is changing how people work. We see it every single day. So what are some of the most common use cases that independent software vendors are solving for right now?
Cedric Courteix (01:52): Yeah, so basically the way we see it at Lenovo, we have these 3Ps, protected, productive, and personalized. This helps us identify scenario where the customer really leverages the device to the bigger extent with AI. So to give you examples, maybe in the productivity, so we have workflow automation, so anything that you do repetitively can be explored by AI and further automated, but it gives you a super boost in your day life. And in the same line of sites, we'll explore one specific ISV that is looking at accessibility, accessibility to the application, but in the sense that we lose a lot of time basically moving from one application to another. And by using this technology, you can leverage AI, the generative AI type of activity, like a Chat GPT, but you are doing it locally on your device. And that's a lot of power for the end user.
Cedric Courteix (03:11): Security is maybe more obvious. We want to protect the endpoint, we want to defend from the people that are doing the wrong things. But because those device have AI capabilities, they can offload a lot of this activity without consuming tons of power and battery. So you are saving on your device in terms of battery life, a good solution would be better than that comes to mind. And finally, personalized AI is an interesting beast. You can see it as a collaborative mind, but if you look at yourself and the type of activities that you do every day, then you can find some optimizations that are really personalized to your individual scenario. So that's the type of applications again we are searching for to fix those type of use case.
Alexa Carlin (04:14): What makes AI PCs uniquely suited to run these ISV apps?
Cedric Courteix (04:19): So they are purposefully built. They have this AI PC core ultra processor that powers this AI workload. So the first element that you see in those devices that you don't see in others is the neural processing unit. The NPU helps you with your intelligent workload, so AI workload, the same way the GPU, the graphical processing unit helps you with graphical application. So it's doing parallel processing but purposefully for AI. And the Intel has a great platform for that and with Lenovo, so we are collaborating to really leverage the framework that Intel has built. It's called OpenVINO that helps us communicate with this hardware component from the application to the devices. So this collaboration is very unique and it's producing great result in terms of battery saving, but also providing the right experience for the end users. If you don't have that type of capabilities, your application will take 10 times and it can basically be very long to give you the result. And when you are in real time asking a chatbot, you don't want to be waiting.
Alexa Carlin (05:56): So if you have the best software now, you also need to pair it with really good hardware.
Cedric Courteix (06:00): Yes, exactly.
Alexa Carlin (06:02): So let's dig into some of these experiences that ISV applications are helping to deliver. What are some of the coolest, most personalized experiences that you've seen so far?
Cedric Courteix (06:14): So I want to talk about a company called Cephable. The owner of that company had a brother with disability, and I think it started this company based on providing capabilities of accessing a device not just on the keyboard, but by gesturing or talking to the device instead of typing on a keyboard like we do traditionally.
Alexa Carlin (06:46): Wow, that's amazing.
Cedric Courteix (06:47): And yeah, it's touching and it's really something that it took a heart, but nowadays, this company has evolved into doing GenAI and being able to do it in your application. So what they've learned by doing the accessibility workload was basically all the controls that are in the application and now they are using it with AI to help you be present in the context of your application and leverage GenAI for translation for rewriting your content or generating new content.
Alexa Carlin (07:27): So making it very inclusive.
Cedric Courteix (07:29): Definitely, because you could use just your voice and be capable of enabling those features. So yeah, that's really a technology that amazed me. And on top of that, they are really nice people behind that.
Alexa Carlin (07:45): And it's part of their heart what's important.
Cedric Courteix (07:49): Yeah, they are really showing it in many fashions.
Alexa Carlin (07:54): What are some of the exciting use cases when it comes to enhancing efficiency for complex multistep tasks like you've been speaking?
Cedric Courteix (08:03): Yeah, yeah. So that would be maybe the case of the workflow that I was talking about. You'll see there's a big buzz in the industry. We call it Agentic AI and the closest I got to that so far is with a company called Iterate. And their product is called Generate AI PC, it has the name in it, but Generate helps you not only do the general chatbot LLM function. So LLM is the large language model. That's what support all these questioning that you are doing with the chatbot, but it's questioning not in the void, it's questioning your documentation. So you are feeding the regular file that you use and the system, it's called retrieval augmented generation that can answer your question based on your data. So it's personal in the sense that nobody else maybe is using the same PowerPoint or the same documents and it can give you really the key element that you're looking for based on your own data.
Alexa Carlin (09:27): Well, summarizing documents is fairly common among use case for AI. So what makes Iterate unique?
Cedric Courteix (09:36): So that's where I was going with the workflow. They allow you to create a tile in the application. So for example, if you are having multiple P&L to compare on a monthly basis, you just feed those two files to the system and it's going to be able to iterate the same steps that you would typically do manually at the end of the month and repeat that as many times as you need.
Alexa Carlin (10:09): So that's like Agentic AI?
Cedric Courteix (10:11): Yeah, yeah. That's where we are getting the Agentic AI, the true promise is that it can find and answer. So you feed them the problem and it'll find the right agent to answer your question. So it might ask multiple agents to figure out the problem a little bit like a mathematic problem. You have a description of what you want, but the system will be able to think about which agent, if it's an addition, subtraction or division. But it's at a much more complex level in terms of figuring out by itself the solution. And I think it's a big promise for AI. We are working toward that, but by increment at this point. Yeah.
Alexa Carlin (11:04): Incremental.
Cedric Courteix (11:05): Because it's more like a single agent that I was describing versus multi-agent collaborator.
Alexa Carlin (11:12): But we're getting that and these PCs will be able to run that.
Cedric Courteix (11:16): Yeah, that's where I see a lot of promises. The first thing we see is the performance out of those devices, but I feel like it's also unlocking scenarios that we haven't encountered yet. And by doing that, it's augmenting my ability, your ability, everybody's capabilities. Perfect.
Alexa Carlin (11:39): Well, so endless possibilities.
Cedric Courteix (11:42): That's the way I view it.
Alexa Carlin (11:43): Unlocking the future. We know security remains top of mind for business leaders today, and AI introduces a whole new element into the equation. How can ISV solutions help users be more protected when implementing AI?
Cedric Courteix (12:01): So what you want to do with this type of workload is protect the large language model, the foundation that is behind the activity of the AI workload. And Dynamo Guard is uniquely positioned because it can really act on your device to control whether it's a large language model that runs on your device or in the cloud, it's going to leverage the local resources to do that. So the neural processing unit that I mentioned before.
Alexa Carlin (12:35): So with that, what are some of the risks of inappropriate Gen AI use?
Cedric Courteix (12:42): Yeah, so there's many risks that we can imagine in the sense that when you are entering your input to AI, you might be asking for the wrong thing. So you want to be able to control, like a firewall will control the device. You want to control what goes into the large language model. The example that I can take is if I'm asking for another employee's salary inside Lenovo, the system should not grant me because I'm not part of the HR, I don't have the ability to do that. So we don't want that the LLM, just because it's capable of looking for this information, go fetch the information that is not allowed for a certain...
Alexa Carlin (13:39): Yeah, I would say that's a big risk.
Cedric Courteix (13:40): That's a big risk. The same thing can be viewed on the outcome. If the LLM is providing something that is not basically suited for the audience, we don't want to allow that. And so there's control that needs to be applied at the input level and at the output.
Alexa Carlin (14:06): Right. It's like making sure it understands proprietary information based on the user that's-
Cedric Courteix (14:12): Exactly, and you are touching on HIPAA regulation for example, in healthcare, GDPR more generally applied to all transaction and business between companies. So it's very important. And as we progress, we know that AI regulations are coming. We've seen that with EMEA or the European regulation, European Act on AI. And what's unique about DynamoAI and DynamoGuard, their solution, they already have the template pre-built to protect you from or to apply those regulations as we go and revisit them as they change also.
Alexa Carlin (14:59): So is this solution unique to Lenovo?
Cedric Courteix (15:01): Yes, no other OEM is partnering with Dynamo AI and we feel really proud of that relationship around DynamoGuard.
Alexa Carlin (15:13): I have one final question. So what makes the partnership between Lenovo and Intel so unique?
Cedric Courteix (15:17): Yeah, I think it's our collaboration. I was on the phone just before coming to this meeting with them. We are working closely to not only find the obvious, but also invent and co-create or develop new applications, whether it's first party or with third party. And that's really where I see a lot of synergies and it starts with the device and we are building the solution of the future for our customers.
Alexa Carlin (15:48): I love that. Solutions of the future.
Cedric Courteix (15:50): Let's see.
Alexa Carlin (15:51): Yeah. Well thank you so much. This conversation has been great.
Cedric Courteix (15:54): Thank you, Alexa, thank you for your time.
Alexa Carlin (15:57): I'd like to thank my guest, Cedric Courteix, Senior Business Development and ISV Engagement Manager, Commercial AI PC at Lenovo for stopping by and talking with us today. And thank you for watching. Visit us online to learn more about how Lenovo can help you accelerate your AI journey on the road to smarter AI for all.
MUSIC: Lenovo, Lenovo.
