Accelerated software delivery with AI

National Veterinary Associates (NVA) is one of the largest veterinary services organizations in North America. With 25,000 employees at more than 1,300 premier locations across the United States and Canada, NVA delivers best-in-class medicine and health services to millions of animals each year. 

Valtech partnered with NVA to test a real-world hypothesis: Could AI facilitate accelerated software delivery without adding risk or compromising quality?

NVA leaned into the opportunity, bringing curiosity and openness to experimentation. Together, we trialed a fully AI-integrated model across the software development lifecycle, from gathering requirements and prototyping through to design and development. 

We embedded AI into every phase of the software delivery lifecycle.

This wasn’t about swapping humans for machines but restructuring the work. Instead of moving work through rigid, step-by-step phases, we built a flow where everything happened in parallel. Progress in one area directly pushed progress in the others.

With AI in the mix, requirements, designs and code weren’t separate documents anymore. They became one and the same. No handoffs, no waiting. Every function was moving forward together. 

• ChatGPT synthesized research, interviews and stakeholder input into usable assets within hours. It produced personas, user stories and journey maps that gave teams an immediate starting point. 

• V0 turned plain-text requirements into interactive prototypes in minutes. Instead of static mockups, stakeholders tested clickable flows almost immediately. This compressed iteration cycles, surfaced assumptions early and reduced downstream rework. 

• Cursor acted like a junior developer, generating React and TypeScript components at speed. Engineers guided it with scope and review, maintaining quality while eliminating boilerplate and accelerating production code. 

By letting AI generate strong first drafts, teams didn’t have to start from zero. That gave them more time to apply their judgment, creativity and expertise.

When orchestrated and guided by people, these tools created a compounding acceleration effect: faster cycles, leaner teams and real outcomes under real constraints. 

We cut delivery time by more than 85%. While a separate, non-AI enabled team estimated it would take seven people 21 weeks to build a production-ready solution, this team delivered a near-production-ready version in just three weeks with a lean team of three experts. 

That’s an order-of-magnitude leap in accelerated software delivery. 

Agile delivery models are being revisited for the age of AI. At Valtech, we’re leading this shift with teams already equipped to deliver in an AI-accelerated world. We continuously explore AI tools as they come to market and consider how they can apply to the work we do with our clients. 

Our pilot proved that AI can accelerate delivery, but it also revealed two truths that will define the future of how we build: 

1. Human expertise is irreplaceable. AI can generate at speed, but it takes people to make the work meaningful. Clinicians brought accuracy, designers brought empathy, engineers brought intent. Together, they turned raw output into real solutions.
   
   
2. Strong foundations unlock everything. Acceleration only works when teams have clear vision, aligned ownership and strong processes. AI exposes any such foundational issues, making organizational clarity as important as the tools themselves. 

In every industry, the organizations that have moved beyond experiments with AI tools must now reimagine how they deliver, at speed and at scale, with clarity and confidence. Explore what that could look like on our Data & AI Revolution page.