‘It’s what we use to basically integrate people, processes and technology to help enterprises operationalize AI at scale, but also with speed and confidence, to show measurable results and let them create repeatable solutions,’ Angela Daniels, DXC’s CTO, Americas, tells CRN.
Global technology services provider DXC Tuesday unveiled a new framework it uses for implementing AI internally and for its clients.
That framework, called Xponential, is DXC’s next-generation AI orchestration blueprint, said Angela Daniels (pictured), CTO, Americas, for consulting and engineering services for the Ashfield, Va.-based company.
“It’s what we use to basically integrate people, processes and technology to help enterprises operationalize AI at scale, but also with speed and confidence, to show measurable results and let them create repeatable solutions,” Daniels told CRN.
[Related: DXC, 7AI Partner On Agentic AI Security Operations Service]
DXC’s approach to operationalizing AI is differentiated in that it ensures the company includes all the people, processes and technology that impact a project, Daniels said.
Xponential includes five pillars related to the implementation of IT solutions, Daniels said.
- Insight: “We’re making sure that we’re embedding governance and compliance and observability to ensure that we’re implementing with responsible AI,” she said. “It’s explainable, and we understand what’s being done by the human and what’s being done by the agent. That’s our key foundation.”
- Accelerators: “These are purpose-built DXC IP as well as partner solutions we bring in in order to accelerate across both the SDLC [software development life cycle] or across a business solution,” she said. “We also have accelerators in things like security or infrastructure.”
- Automation: “Think about the many agents that we pull together that we have defined across agentic frameworks so that we can focus on an outcome,” Daniels said. “This allows us to optimize across whatever processes that we may be implementing and provides a level of continuous learning so it always gets better.”
- Approach: “The human stays in the loop,” she said. “AI is used to amplify human expertise. Always keeping humans in the loop is important to us.”
- Process: “It’s how we are approaching it,” she said. “We start small, and after we are successful and achieve some early wins, then we stack those to get velocity and scale. That’s the repeatable part we talked about.”
While these five pillars are important to developing AI solutions, they are not something all AI developers always do, Daniels said.
“It’s the way we bring them together across those five pillars that differentiates us,” she said.
As an example, Daniels said a challenge clients face is that if DXC brings in AI and related accelerators, how will they be able to see the benefit?
“For instance, what productivity is it giving me across the SDLC?” she said. “Within this blueprint we have both the tooling and the approach to understand how a developer will be accelerated across the SDLC and be really precise about what accelerator to bring, when to bring it and why. That, I would say, is the key difference there.”
An important benefit of the Xponential blueprints also is repeatability, Daniels said.
“Xponential functions compound over time,” she said. “So we’re able to go in with the customer, define the issue, and start small across a couple of use cases. And from that, that client can then take what we’ve done in that one area and then continue to apply it across others. When I’m talking to clients, I often give this example. Let’s say I wanted to purchase a new car, and I’m promised that it’s going to be 30 percent faster. It’ll get me to work 30 percent faster. Now I have 30 percent of my time I can use and get value from by doing other things. Maybe I stopped by Starbucks, or maybe I’m able to start work earlier. Those are the things. [That’s] new time that it’s created. It’s not just about speed, it’s about what that speed enables, and that’s what Xponential delivers.”
In a more concrete example, Daniels cited Singapore General Hospital where DXC helped develop a solution called AI2D, or Augmented Intelligence in Infectious Diseases, Daniels said.
“The solution wasn’t just about improving antibiotics,” she said. “We use augmented intelligence and were able to improve antibiotic decisions, and that was able to help reduce unnecessary antibiotic usage. … It’s not just about antibiotic precision. It’s freeing up physician time or accelerating decision-making, and it’s creating this reusable AI governance model that now other departments can adopt. Think about that 30 percent value I talked about in the car example. For Singapore General Hospital, we’re taking that value to free up time for innovation.”
Looking ahead, DXC is embedding AI in many of its offerings and wants to include the Xponential blueprint for combining AI and human agents, Daniels said.
“Last year, we made all of our engineers go through AI conversant training so they understand how do I implement prompts? How do I think about it from an agentic perspective?” she said. “And now they are moving to understand how to leverage agentic software engineers. ChatGPT, in its current form, is really an AI assistant. You go and ask it a question; it gives you a response. AI augmented is more involved with orchestrating all of these agents to complete my development task. And so we’ll continue to double down on enabling our teams to be more human-plus and, of course, building accelerators and automation there that our clients can leverage.”