AWS Summit Washington, DC 2025 | Valerie Henderson, Caylent

Clips
News
More from AWS Summit Washington, DC 2025

Valerie Henderson

President

Caylent

Inside the AI shift in government: 11 perspectives on public sector ops transformation

Public sector transformation is entering a critical phase, driven by the rise of agentic AI and digital labor that doesn’t just automate, it adapts. But the biggest obstacle isn’t the technology. It’s the systems, silos and culture built for a slower, pre-AI world.TheCUBE’s John Furrier discusses public sector transformation.What’s really holding back progress in the public sector isn’t a lack of innovation — it’s the difficulty of adoption. From aging infrastructure to deeply rooted bureaucracies, systemic inertia remains the biggest obstacle. Meanwhile, AI is rapidly reshaping the role of government, raising the stakes for timely transformation, according to theCUBE’s John Furrier

play_circle_outline AWS DC Summit and cloud advancements in public sector and commercial sectors

play_circle_outline Challenges and blockers in data privacy, regulatory compliance, and trust factors

play_circle_outline Impact of robotics and AI integration in healthcare and life sciences

play_circle_outline Transition from generative AI to agentic AI for decision-making in healthcare

play_circle_outline Focus on deep domain expertise and AI frameworks to deliver results for customers

Info
Transcript

Valerie Henderson, Caylent

Valerie Henderson

President Caylent

Valerie Henderson, president of Caylent Inc., joins theCUBE’s John Furrier and Scott Hebner at the AWS Summit Washington, DC event to discuss how generative AI and cloud infrastructure are driving transformation across industries. As the leader of an AWS premier partner, Henderson shares how Caylent delivers cloud-native solutions that support growth in data modernization, AI and app transformation.

The conversation explores how AWS services — particularly Bedrock — help address challenges around data privacy, compliance and automation in fields such a... Read more

explore Keep Exploring

What is the speaker's take on the current moment in time regarding the advancements in technology and data utilization in various industries? add

What are some important considerations when it comes to security and compliance in healthcare data management? add

What are some applications of robotics and AI coming together? add

What potential benefits could agentic AI bring to the healthcare industry beyond just generative AI? add

What framework do you use to accelerate the migration of legacy databases to AWS? add

bolt Powered by CUBE AI

Valerie Henderson, Caylent

search

>> Welcome back, everyone. We are here on the show floor at AWS DC Summit, formerly the Public Sector Summit. As it all comes together, I'm John Furrier, your host with Scott Hebner, research analyst at theCUBE. Valerie Henderson, president of Caylent is here, CUBE alumni. Great to see you again. You look fabulous.

Valerie Henderson

>> Yeah, great to see you. Thank you. Thank you for having me.

>> So welcome back. So this show is really all about how public sector and commercials come together, cloud going next level with gen AI. Massive tsunami of apps coming that are going to be agentic, the data layers. We were just at the Snowflake Summit. I'm flying out tonight to go to Databricks. A lot of action happening.

Valerie Henderson

>> Yes.

>> So this is going to create a lot more value pretty quickly. You're in the middle of it.

Valerie Henderson

>> Absolutely.

>> What's your take of this moment in time?

Valerie Henderson

>> I had a mentor that on his wall, had the printing press and the telephone, and it truly is the next golden age. I really do believe that. People can't even comprehend what the next year, two, three years are going to look like. It's going to be revolutionary. It's not a hype cycle. It isn't. And we're really seeing it. We're seeing healthcare organizations put it into production, not just POC.

>> Valerie, how about your business? Where are you guys now? Give us some stats, a quick overview.

Valerie Henderson

>> Yeah, it's been amazing. So we're a premier AWS partner. We're actually all in with AWS, so we do have an opinion that AWS is the best hyperscaler to build on. I really do believe that. We've had incredible growth. It's been 10X over the last three and a half years, a wild ride. We really have concentrated on being extremely strong in the level of our technical depth within AWS, generative AI, data and application modernization, all being underpinned by of course, migrations because you know, that is still a need in this space.

>> Yeah. And the tooling right now and some of the coding-

Valerie Henderson

>> Oh, my gosh....

>> has been phenomenal. So sales and marketing and coding are the two hot areas. You're seeing that boom. What is that doing for you and your clients in terms of their adoption from a lean in standpoint, impact, POCs? Can you share?

Valerie Henderson

>> Yes, absolutely. There was a period of time when everybody's boards just said, "Do something," and now we're actually seeing these experiments of POCs turn into real production workloads, integrated into people's product. We work with a lot of different ISVs and we're not talking a chat bot. We're talking generative AI, UI and engagement to really make you feel a different way when you're using these products, which is really special.

Scott Hebner

>> One of the things I've been seeing in the research that I've been doing and talking to dozens of companies is the aspirations are there, a lot of activity going on, but there are some things that are some headwinds.

Valerie Henderson

>> Blockers.

Scott Hebner

>> Blockers like the trust factor in terms of data privacy and regulatory compliance. And especially when you get into agents, if you're going to have these things to actually help you make decisions, you got to trust it, which means it needs to be explainable. But let's start on the data layer, right?

Valerie Henderson

>> It's everything.

Scott Hebner

>> It's everything, right? Privacy and regulations.

Valerie Henderson

>> Of course.

Scott Hebner

>> Especially in healthcare, right?

Valerie Henderson

>> Massive. We were actually talking about that about an hour ago. The thing that's special about AWS is within Bedrock in terms of their level of security and compliance, it literally covers all bases. They also have guardrails that really help make sure that you have clarity on hallucinations and what is and is not factual and real, but one of the biggest blockers in healthcare is what you said, security and then a fear of the unknown or privacy or mistakes. And what I would tell people is work with a partner that has a lot of experience doing that. It does not have to be Caylent, but there are hundreds of thousands of partners out there that have a lot of experience. Because that's the other gap is not having the knowledge or understanding of how to use these-

Scott Hebner

>> Yeah. The best practices.

Valerie Henderson

>> Yes. Because what it is, and you must have really strong evals. Like in any software development, if you do not have strong evals and foundations, nothing sticks. So that is really important from the beginning.

>> Financial services obviously always an early adopter, check. Put them aside for a minute.

Valerie Henderson

>> For sure.

>> Healthcare and life sciences are two hot areas. One, they need supercomputing capability, which now the cloud is becoming with all the great infrastructure performances. But between Bedrock and SageMaker, you now have an under-the-hood with SageMaker capability, and with Bedrock you have optionality on models. Talk about the impact that means to you guys. Because as you roll out and prototype and deploy to production, what does that do? Faster speeds? Can you share your thoughts on this? Because I think this is where the action is moving very quickly to.

Valerie Henderson

>> Well, number one, to your point, Bedrock, you have all the optionality and flexibility. We talk about it, the Curacao model, you might implement a model and that model is different within two, three months. So being able to leverage Bedrock and have those evals where you can upgrade those models, real-time is massive. But specifically in healthcare and life sciences, there is all of this imaging data. We work with a company called VITA, and they are specifically focused on vascular imaging to be able to help with diagnostics and drug testing. It is very difficult to take all of those images and actually de-identify them because of privacy concerns. And we've been able to leverage AWS technology like Rekognition and Textract and Bedrock to actually build almost a co-pilot to increase the viability and the speed at which they get these results. Which means you go from potentially years to months on getting these drug trials testing done. And I read recently it can be anywhere between $600,000 to $8 million a day when you're testing a drug and testing its viability, so every minute counts. So it's massive to see this type of change.

>> Healthcare and life science have a common thread. Robotics are a big part of it. So robotics and AI are coming together. We hear in all the keynotes, robots, we were showing some videos of robots doing some things, shooting guns. We think they'll be calling balls and strikes in the baseball game.

Valerie Henderson

>> I first a dish washing robot, like, sign me up. I'm in. Ironing, vacuuming, I'm in.

>> It's coming quickly. I missed that boat. I wish that happened when I had my kids. But robotics has a natural affinity towards machine learning and software, hardware relationship. Healthcare, a lot of robotics, life sciences, robotics, multi-purpose AI is a huge thing, there's a lot of leverage.

Valerie Henderson

>> Yeah.

>> Can you share your thoughts on that? Because we're seeing a lot of customers rethink their deployments with that in mind, because it's not a bespoke IT project.

Valerie Henderson

>> No, not at all. We actually were just on a panel and a gentleman was talking about, he went into a lab and they actually had... This used to be a lab where it was all bench scientists and it's now split in half with all robotic automation, and then half of the room is bench scientists. And he said, "Okay, what is this going to look like in 10 years?" And he's like, "There probably will be one bench scientist and the rest will be robotic automation." And I think just thinking about that at the top of the design is the biggest part. You have to be able to expand and think bigger, otherwise you are going to design things too close and not being able to take advantage of this.

>> And Bedrock fits in beautifully because you have an LLM plug into the robotics-

Valerie Henderson

>> That's right....

>> and be purpose built. So customers are going to want their own language models.

Valerie Henderson

>> Absolutely.

>> And interface with the big ones, distillation's a big factor.

Valerie Henderson

>> To refine and make sure that it is as relevant for that specific use case as possible. It's not about the model. And that's why I think Amazon was so smart with the way that they designed.

Scott Hebner

>> Right, right. Because I think LLMs are just a building block to be more specialized models, domain specific. You need semantic reasoning, you need causal reasoning. These things are all going to come together over time. I think that's the big difference between a genitive AI and agentic, right?

Valerie Henderson

>> Yes. Yeah.

Scott Hebner

>> Decision making. You go from predictions to judgment. So I'm curious to what you're seeing, and especially in healthcare, the notion of agents, digital labor, actual decision making, whether it's human in the loop or not, changes the game again, doesn't it?

Valerie Henderson

>> Yeah. Somebody said the best analogy is generative AI is a football coach on the sidelines, and agentic AI is a literal quarterback that has some level of automation involved. And I think that that's true. If you think about the level of resistance that exists within healthcare right now of just generative AI, I want people to expand the thinking beyond just that because what you can do behind the scenes with agents is incredible. And nurses, for instance, spend half of their week doing administrative tasks. And if you think about re-purposing that time to interfacing with patients, which is why they got into doing what they do, and having agents work in the background, connecting those workflows, that's going to change the face of healthcare. Think about how many people hate going to the doctor.

Scott Hebner

>> Right.

>> Yeah.

Valerie Henderson

>> Because it's not always-

Scott Hebner

>> They you that sheet.

>> Well, also sometimes if they just take notes wrong, you don't get the referral.

Valerie Henderson

>> Exactly.

>> Little things like that-

Valerie Henderson

>> It's crazy....

>> get caught up in the human error.

Scott Hebner

>> Well, the biggest thing is you got to manage your own healthcare because if you have several different kind of doctors involved, they never talk to each other.

Valerie Henderson

>> That's right.

Scott Hebner

>> And so yeah, agentic AI has the possibility to bring-

Valerie Henderson

>> To bring that together.

Scott Hebner

>> Yeah. The innovation cycles are moving so rapid. It's like-

Valerie Henderson

>> It's crazy....

Scott Hebner

>> you can't fall behind now.

Valerie Henderson

>> No, you can't.

>> Computer vision.

Valerie Henderson

>> Oh, yes.

>> Multimodal.

Valerie Henderson

>> Yeah.

>> We got the LLMs now. You got computer vision, healthcare, imagery.

Valerie Henderson

>> And AI, all of it. Yeah.

>> Take us through a day in the life of what you guys do. Because it's technical, you guys have deep domain expertise on AWS going all in. You mentioned that. The customers have a change management problem culturally. How do they get up and running? Clinicians, the humans on their side, back office, you have those three areas. What's your thoughts on that?

Valerie Henderson

>> We actually built a OCM change management business because of generative AI. Cloud needed it at some point now, but we are in a place now where people understand it. They might not have the expertise, but they get it, but generative AI needs so much help. Our theory is four kind of sided. Number one, you have to co-develop. Number two, you have to have tight feedback loops. Three, you need very clear earned trust milestones. When this happens, when this measurement gets to this point, we'll know it's working. And four, you need reinforcement. So how do you train, how do you incentivize? That's huge. People don't think about that.

>> Scott and I were talking about this, is that service business could be any vertical.

Valerie Henderson

>> Yeah.

>> Research analysts, ours, media, the service leverage you get with operating with the cloud on those four points is critical. And you can almost take your last four points and make them boilerplate.

Valerie Henderson

>> Totally.

>> If you do these things, it's a process, it's not a transactional deal.

Valerie Henderson

>> That's right.

>> How would you educate someone on that? Because not everyone sees that. How would you simplify that value proposition?

Valerie Henderson

>> Totally. Well, I typically, I will sit with somebody. And what we like to do is we like to meet with line of business. And I actually want to walk through the entire problem statement with them. What does that look and feel like? You have to prompt them in the same way you're prompting generative AI. You have to be really thoughtful about the questions that you're asking, and ultimately you can lead them to where they are outlining those four specific points of what's needed within their organization.

Scott Hebner

>> Yeah, I going to say, they come up with it, right?

Valerie Henderson

>> Yes.

Scott Hebner

>> Yeah.

Valerie Henderson

>> Yeah.

>> So here's my question to you. As a leader, you said you made a few bets. I'm connecting the dots here. You bet on AWS.

Valerie Henderson

>> Yep.

>> You saw the gen AI wave.

Valerie Henderson

>> Yeah.

>> How did you instruct your team to do things differently and what did you guys intentionally do from a strategy standpoint that was turned into execution? More leverage with the tech?

Valerie Henderson

>> Yeah.

>> More end-to-end relationship with the customer, probing for problems, reinforcements, things you mentioned. What were the key things of your... Because you guys are growing, so I think it's a good template.

Valerie Henderson

>> Yes. Yeah, yeah. I'm grateful and I never take that for granted. We're disciples of, only the paranoid survive. You always need to be looking around corners, but I think the next wave is already what we've talked about, is agentic. But in our services space, it is how we are leveraging AI to deliver results for customers. And that's something that I think is not necessarily being caught everywhere because it's really disrupting our own business. But how are we using AI and AI frameworks to deliver results faster? Because prior, it was headcount, revenue, hours. Now it needs to be-

>> What do you guys do? What specifically do you guys do?

Valerie Henderson

>> So we actually have what we call our accelerate framework. It is applied at whatever you want from a backend infrastructure standpoint, but more commonly, database right now. There are so many legacy databases that are just sitting there and they hadn't been moved because the cost and the time. And we are shaving almost 70% off this time. So we can go back to customers and say, "Hey, you didn't want to touch this. Now you can, and it can be AWS native and it can be more integrated into what you need at the data layer to take advantage of generative AI." I think that that's a huge part of our strategy.

>> That moves the needle

Valerie Henderson

>> Massive.

>> The engagement goes to the next level, and those four things you mentioned kick in, and then your motion has changed-

Valerie Henderson

>> Significantly....

>> with the customer.

Valerie Henderson

>> Yeah, absolutely. The demand is different. And I always say, you have to be ahead of where the customer is going.

Scott Hebner

>> That unlocking the value of your data is so important. Some of the research we recently did, today, only on average, about 40% of data is actually used-

Valerie Henderson

>> It's crazy....

Scott Hebner

>> by a business. It's going to go down to 6 or 7% over the next three years, unless we do something different. And I think what you just said is something different. You got to use AI. You don't necessarily have to move data, load it up in the Mac trucks, ship it up to the cloud. But there's new ways to get to all that data, to assimilate it, to ensure, we were talking before about the privacy and the compliance, and you got to start thinking about that today because your data is your biggest asset.

Valerie Henderson

>> It is. People have to be thinking about not being IT, it's IP and how you're getting the value out of your data.

>> So modernization, cloud native on AWS, value-oriented motion.

Valerie Henderson

>> Big time.

>> You're now in a position. What's next? What are the key things you're focused on?

Valerie Henderson

>> I think we are going to see a need for deep level of domain expertise as it applies to agents. So for us, having this partnership with AWS within specific industries and verticals is so key. For instance, higher ed has a set of platforms that do not talk to each other and they need to talk to each other. Can you build in agents and help build a wrapper around that for a higher education firm? You absolutely can.

Scott Hebner

>> Let me ask you just in assimilation of all the conversations I've had, like dozens and dozens over the years. Three keys here from what I've learned. One is you got to focus on nuanced markets or nuanced skills because if it's about people making... I'm talking about agentic, right? About making judgments better, making decisions, planning, think about all the niches there are for very nuanced skills in the industry. It's uncountable.

Valerie Henderson

>> It's infinite.

Scott Hebner

>> Right. Two is target labor spend, which is the majority of the spend of most businesses, not just IT spend.

Valerie Henderson

>> That's right.

Scott Hebner

>> And then thirdly, rethink software bottoms up. It's not just SaaS modeling.

Valerie Henderson

>> No.

Scott Hebner

>> And those are three keys to being successful. And it sounds like you guys are actually another example of doing all three of those. True?

Valerie Henderson

>> It is uncomfortable to disrupt your own business. And even being a firm of our size, you can feel sometimes resistance, but I'm really grateful that we've got a team that is young and excited to be tackling this.

>> Yeah. Valerie, we've been talking on theCUBE here and other events. This comes up a lot. It came up last week in Seattle. Some industries already have better AI readiness and they don't even know it, like healthcare and some other ones because of compliance. They've been tagging data for a lot reasons, but now they're in a perfect position. Others haven't. Are there certain areas, verticals that are more prime right now for gen AI? You've mentioned healthcare. What other ones do you like that are hot and which ones need to get their and stuff together?

Valerie Henderson

>> Yeah. Yeah. Media we work with quite a bit and there's a lot of opportunity there. But I will tell you, manufacturing is an area that I think is, albeit not sexy, one that has a ton of unlock from a data standpoint. We worked with a company called BrainBox that was acquired by Trane that did HVAC based intelligent generative AI. We worked with a company called Simmons, that's a plumbing manufacturing organization that builds generative AI co-pilot to tag waste issues with water. They're thinking they're going to save 5 billion gallons of water. And I think that looking at those areas where there are devices that are spinning out data that you can leverage and actually monetize and empower and/or create some level of sustainability.

>> Another conversation I'd love to get your thoughts on, you're like a co-host now on theCUBE. just the expert. They have value creation and value extraction always are the two sides of the coin. So they create value, we see that happening a lot. And for the reasons you mentioned, get some smart people. They could be an individual, could be a group. They build a system, as a systems thinking. The value extraction is where the customers are now first generation, like, "Wow, I didn't know I could save five gallons." What's the value? How do I value that? How do I value the project?

Valerie Henderson

>> Oh, my gosh.

>> Not a TCO calculation on cost. Is it now revenue or other surprise impacts that are beautiful?

Valerie Henderson

>> That's right. Yep.

>> What is that extraction, value extraction piece? What's your observation? Can you share any content around that?

Valerie Henderson

>> Yeah. We were with a firm actually specifically that was going through this exercise of leveraging generative AI within HR. And they're like, "How are we monetizing thinking about the value we're getting out of this?" And it did require hours and hours of sitting down with the end users and talking about like, "Okay, how long does that take you? How much are you being paid? How do you actually look at that value of those hours spent?" And I think that that requires a third, fourth, fifth level of questioning and understanding of these unique businesses. And that's the part that I think is going to take time.

Scott Hebner

>> Where this is going to be a really interesting is today it's a measurement about productivity.

Valerie Henderson

>> That's right.

Scott Hebner

>> Because generative AI is helping you do things more, but when you start getting into agents, especially down the road, it may help you make actual decisions. How do you put a value on a better decision? It's a better plan.

Valerie Henderson

>> How do you think about that in a monetary fashion?

Scott Hebner

>> That's big.

Valerie Henderson

>> Absolutely.

Scott Hebner

>> How many wrong decisions does the organization make?

Valerie Henderson

>> We all get decision fatigue.

>> Our theory on this, I'd love to get your reaction on this, is that when people figure out the valuation techniques, SLAs are next.

Valerie Henderson

>> That's absolutely without question, a hundred percent. I think that we talk about that a lot. Time, how quickly should it turn around? I also think the way that our services are being priced, does that look different? Is it outcome-based, token, whatever it might be. So I think all of that is ripe for change.

>> Valerie, great to have you on.

Valerie Henderson

>> Thank you all.

>> Awesome content. What's exciting you these days? What's the coolest thing you're working on?

Valerie Henderson

>> Oh, I would say the coolest thing I'm working on is with a company called NeoGenomics, and it is actually a pathology organization working with oncology. My mom is a six-time cancer survivor, and this is a firm that is actually looking at NGS, so the next-gen sequencing and helping match patterns faster. And when you're in a waiting room, waiting on results for these tests, minutes matter. And so being able to be a part of that, that's going to make that whole experience better. Invaluable.

Scott Hebner

>> That's great.

>> The care, the frequency, the acceleration.

Scott Hebner

>> Oh, my gosh. All of it.

>> And by the way, you don't have to come back.

Valerie Henderson

>> Totally.

>> Go home, come back. Wait a minute. Where's my thing?

Valerie Henderson

>> Exactly.

>> Valerie, thank you for ending the show for us here on theCUBE.

Valerie Henderson

>> Oh, good. Yes. All right.

>> That's a wrap up as well. Thank you for the comments. Again, congratulations.

Valerie Henderson

>> Thank you so much.

>> We'll see you in New York in July.

Valerie Henderson

>> Yeah, absolutely.

>> All right. For Scott Hebner, I'm John Furrier with theCUBE. We are wrapping up AWS DC Summit. Thanks for watching.