AWS Summit DC 2026 | Lillian Chang, SAP NS2

Clips
More from AWS Summit DC 2026

Lillian Chang

SVP of Product Strategy

SAP NS2

play_circle_outline FedRAMP compliance ensures secure, auditable handling of sensitive federal data

play_circle_outline SAP AI Core Turns ERP into Action: Natural-Language Procurement Insights and FedRAMP SAP via SAP NS2 and AWS GovCloud

play_circle_outline From System of Record to System of Action: Embedded AI and SAP AI Core Securing Custom ERP Intelligence

play_circle_outline AI-Powered Procurement: Natural Language Alerts and Scalable Invoice Matching for Supply Chain Resilience

play_circle_outline Fast, Secure AI-Enabled ERP Deployments: Overcoming Legacy Systems and 6–12 Month Authorization Timelines

play_circle_outline SAP's roadmap: embedding AI in line-of-business solutions and supporting custom AI use cases securely

Info
Transcript

Lillian Chang, SAP NS2

Lillian Chang

SVP of Product Strategy SAP NS2

In this interview from AWS Summit Washington, D.C. 2026, Lillian Chang, senior vice president of product strategy at SAP National Security Services, joins theCUBE + NYSE Wired's Gemma Allen to discuss how AI is transforming government ERP systems from passive records into active, mission-ready tools for federal agencies. Chang explains how SAP NS2's decade-long partnership with AWS GovCloud provides a secure, FedRAMP-compliant foundation for agencies looking to put AI to work on core enterprise data. She details how SAP AI Core, running on AWS GovUS, removes ... Read more

explore Keep Exploring

How are you working with AWS GovCloud and SAP’s AI Core to provide secure, FedRAMP-compliant AI solutions for federal customers? add

How is SAP transforming its ERP from a system of record into a system of action using AI (e.g., SAP AI Core), and how do SAP, SAP NS2, and AWS share responsibility for FedRAMP compliance and security for government customers? add

How does SAP AI Core enable government customers to run secure, auditable natural-language queries on ERP procurement data (for example, to assess supplier performance and purchase risk)? add

How can AI be used to improve supply chain resilience (e.g., preventing stockouts via alerts and natural-language queries) and to automate procurement tasks such as three-way matching of purchase orders, goods receipts, and invoices? add

What is the FedRAMP authorization process for AI solutions, how long does it typically take, what must an authorizing official do, and how can using FedRAMP-authorized services (e.g., AWS and SAP NS2) simplify that? add

What is SAP's roadmap for embedding AI into its line-of-business solutions, how will you support both embedded and custom AI use cases, and what is your role in securely delivering those solutions to customers? add

bolt Powered by CUBE AI

Lillian Chang, SAP NS2

search

Gemma Allen

>> Welcome back to theCUBE's live coverage here on the ground in Washington DC for AWS Summit 2026. We had a jam packed day yesterday. We talked all things AI and the future of government and public sector, and today we have an incredible lineup. We are going to talk about cybersecurity, AI and ERPs, how AI is fundamentally shifting the pace of innovation at government. Joining me now for a conversation on the data that feeds that AI, and how important it is that that data flows in a very secure fashion, is Lillian Chang from SAP NS2. Welcome, Lillian.

Lillian Chang

>> Thank you very much, Gemma. Thank you for inviting me.

Gemma Allen

>> Well, we're delighted to have you. So what an interesting time. First of all, how are you feeling? It's been a great two days so far.

Lillian Chang

>> This is an amazing conference. Every year I attend, it grows bigger and bigger. I think I heard 26,000 attendees.

Gemma Allen

>> Wow.

Lillian Chang

>> That is amazing.

Gemma Allen

>> Wow. And what a week in Washington DC, right?

Lillian Chang

>> Yes.

Gemma Allen

>> It's so historic.

Lillian Chang

>> Yes, it is.

Gemma Allen

>> So speaking of history, and making history, and the continuation of this great country, your role, Lillian, is very interesting in that you to SAP and S2, the security arm of SAP, feed government systems with very important data that is used for mission-critical tasks, right?

Lillian Chang

>> Yes.

Gemma Allen

>> We're talking making sure military equipment gets there on time.

Lillian Chang

>> Correct.

Gemma Allen

>> All sorts of important things that need to happen that we sometimes don't really hear about or think about.

Lillian Chang

>> Right.

Gemma Allen

>> I think AI has really changed a lot of the narratives in a lot of spaces, and it's interesting because ERP has always been a backend thing, right? Something people can't think about, but it's more important than ever in some ways. Talk to me about what's changing for you and the relationship you have with AWS.

Lillian Chang

>> Yes. I love to. We've been actually working with AWS Gov for over a decade now. So what we do is we build our SAP solutions AI included on AWS gov. And what we do is we go through the whole federal security process of getting it FedRAMP and assessed. And why is this important? So our customers are asking a lot of times with AI, how can I ensure that I'm using AI securely and safely? And now with SAP's AI core on AWS GovUS, we can assure that. And then now what's happening is that pivoting that question to, how do I get to use it right now? So what we're trying to do is make sure that our customers don't have to worry about the security part, and that's our role there. We're going to go through and get all the federal compliances that are required to ensure that federal data that is sensitive that should not be out publicly, is in a controlled, audible, authorized boundary.

Gemma Allen

>> Let's talk about SAP for a second because when I think about SAP, having spent 20 years in tech, I worked with it myself. It is the most solid, it's like a tractor engine. You just know it's going to run. And it's a single source of truth.

Lillian Chang

>> Yes, exactly.

Gemma Allen

>> That data though, sometimes it's difficult to retrieve data from for some folks. You need to know the codes and all that, but it is a powerful, powerful engine. Getting that data though quickly, at the speed of AI seems like for someone who hasn't used SAP in 20 years, a challenge.

Lillian Chang

>> Yes.

Gemma Allen

>> Talk about like what is changing from the perspective of it being just a system of record, to being now a system of action, per se?

Lillian Chang

>> Yeah. So what's happened is, like you said, our core ERP is the business transaction record. So what we're seeing is government customers, what do they have to do? Well, I have a whole bunch of procurement data in there. Well, instead of running reports, I want to do natural language queries on who are my suppliers that are performing, what are my biggest risk with any of my purchases? And so that's what we're seeing. We built SAP AI core to make sure that that custom AI will work on top of SAP ERP data. And that's nice, for our customers, they don't have to worry if they use another model that it's not going to be secure. Because what we also do is make sure that the models, the LLMs are also FedRAMP and will work with our AI solutions. So as a customer, all you have to do is, hey, query your ERP solution with some AI capabilities, and you can assure that it's very secure, and it's audible again and it's accurate. It's a single source of truth and that's what we're trying to do for our customers.

Gemma Allen

>> FedRAMP is an interesting conundrum, isn't it?

Lillian Chang

>> Yes.

Gemma Allen

>> Because it's complicated and intentionally so, right?

Lillian Chang

>> Yeah.

Gemma Allen

>> In terms of, but it provides a level of security that's absolutely a necessity in the space you're in. In terms of the relationship between AWS and SAP for NS2, who owns that FedRAMP relationship? If you're a customer of either of yours, who fundamentally does the buck stop with?

Lillian Chang

>> We both do. So AWS gov, they have the infrastructure. So they ensure they get their FedRAMP solutions. At SAP, we have software solutions, applications, platform. We also do the same at NS2. We make sure we get the FedRAMP at those levels as well. So all of us are responsible for that FedRAMP compliance. SAP NS is unique for SAP. We're all US citizens on US soil. What we do is we operate and maintain SAP solutions, and as well as get those security compliances for our federal government customers.

Gemma Allen

>> Bring me behind the curtain of SAP NS2 for a second. Help me imagine some use cases, and I guess specifically some use cases that might be changing from the perspective of the speed that data is needed to be ingested at. Talk me through how things have changed even in the last five years. Because if you think about it, the pandemic was a very interesting learning for us. It made us realize that data and ERPs, they're a matter of national infrastructure too.

Lillian Chang

>> Yes, exactly. So you bring up the COVID situation with supply chains. Supply chains need to be resilient. I think we learned a lot. What did we see? A lot of outages of inventory. So what's happening with AI with our data, is you're able to do natural language queries and be alerted before these things occur. You don't want to have a stock out. You want to prevent it. You want to focus on your mission and make sure the inventory is there. So AI is able to go through all that data very quickly, alert you. It's also able to do and execute some of those commands versus just doing queries. We're also seeing this in procurement. A federal government agencies vary in size. Some have 10,000 invoices, some have 100,000 invoices. So you can imagine when you have to do a three-way match between a purchase order, a goods receipt and an invoice, that can be manual intensive. What we're seeing with AI is utilizing AI to actually do most of the work automatically. There might be some exceptions, and then as you process those exceptions, the AI's going to learn and then they'll be able to execute. So imagine being able to go through 10,000, 100,000 invoices with a matching with a purchase order.

Gemma Allen

>> It's recursive, I guess. In the best possible way.

Lillian Chang

>> Exactly.

Gemma Allen

>> Talk about some of the challenges, because we know that especially in the space you're in, national security, it's a very locked down environment, right?

Lillian Chang

>> Yes.

Gemma Allen

>> It wasn't designed to be technically agile per se, they don't want to be leading any curve when it comes to tech innovation because they need to make sure things are extremely secure. How is that changing? Talk me through on those challenges that we've all known, like legacy systems, legacy data, systems that are 30, 40 years old that pre-date us, they still exist, right? That is the reality.

Lillian Chang

>> Yes.

Gemma Allen

>> How are things changing, Lillian? Bring it to life a bit.

Lillian Chang

>> So what you're going to see is the FedRAMP process. You just asked about it, the authorization. That can be lengthy. That could take a lot of time. So what we're seeing with our customers, particularly with AI, you have an authority officer who has to say, "Hey, I'm going to assume the risk that this AI solution is secure and I have to make sure data doesn't go out." So if you do that yourself, what you have to do is you have to get your architects involved, you have to get your security compliance involved, and you have to review everything and that process can be actually lengthy. It could be six to 12 months, and that's if you invite everyone early at the table, but you know that's not reality. So this timeline is long. So what we do is, "Hey, we've done that work for you." AWS and SAP N2. We've done our FedRAMP for our infrastructure, for our platform and our software. As an authority officer, you can just go, "Oh, it has FedRAMP. I can utilize it. Now what I can do is focus on my AI use cases, on my mission-critical systems."

Gemma Allen

>> And when we think about the go to market for these partnerships, right? You guys are working in organizations with folks that have probably been in the industry a long time. They have preferences, loyalties, we all build those up. How is that changing from the perspective of this AI chaos that we're somewhat surrounded by? When you guys are marketing this joint solution, are you having the conversations with the same folks you're having conversations about five years ago? Or are those roles are also changing?

Lillian Chang

>> So you're going to see a lot of new roles. I think we heard about data. You were talking about all this massive data. Now there's Chief Data Officers that we work with. They have to make sure the data is clean and accurate. Because if you're using AI scenarios off of data that's not clean, you're going to have bad results. We're also working a lot with the CTO, the CTO conversations about the AI architecture. We have a lot of conversations with CISOs, the security officers, ensuring that we're checking the box for them as well. So that's the trend we're seeing. It's more on the data, technology and security side.

Gemma Allen

>> I guess we're in a world where suddenly everyone is a technologist, right?

Lillian Chang

>> Yes.

Gemma Allen

>> That's the truth. They're expected to be.

Lillian Chang

>> Technology's moving so fast.

Gemma Allen

>> So Lillian, SAP, it's hard to think of a more steady company in a lot of ways. What's ahead for you? What does the next 12 months look like? And from the perspective of building, breaking and innovating, what's the roadmap?

Lillian Chang

>> Our roadmap is, as you said, continuing to support our line of business solutions, whether it's finance, supply chain, human capital management. What we're looking at is embedding that AI. What does SAP know best? We know our business processes. We know the business transaction data. So what we're trying to do is build a lot of those AI use cases within the product. And so that means our customers won't have to build it. So we call that embedded AI. Well, we also know that every customer has unique use cases, so we have a custom AI as well and we want to support that. So my role is really figuring out to make sure we get these securely in SAP NS2, and out to our customers very quickly. Everything is about speed right now and security.

Gemma Allen

>> Yeah, security is certainly everything. Lillian Chang, thank you so much for joining us on theCUBE.

Lillian Chang

>> Well, thank you for having me.

Gemma Allen

>> I'm Gemma Allen. We are here with theCUBE live on the ground at AWS Summit 2026 in Washington DC. The energy is palpable here, no sign of it fading. Don't go anywhere, we've got a great lineup of guests.