Tech Spotlight: Alex Sauer-Budge, Ph.D.‍

We sat down with Alex Sauer-Budge, SAVVI’s CTO & Co-Founder.  He is an MIT-trained leader and one of our many in-house PhDs, to get his perspective on the technology changes sweeping through the benefits industry and what this means for organizations and consumers. 

‍

There's a persistent assumption in the tech industry that ethics and innovation are at odds, that designing for fairness, transparency, and user protection means slowing down. In fintech and benefits technology, we’re witnessing the opposite play out.

When you build with ethics embedded from the start, you build better products. Products that people trust. Products that get used. And in financial guidance, where the whole point is to change behavior, trust isn't a soft value. It's the mechanism by which everything else works.

‍

Q: How do you define "ethical AI" in the context of benefits and financial guidance?

For us, it comes down to four principles: explainability, autonomy, data minimalism, and transparency about incentives.

• Explainability means an employee should understand why they're being recommended a particular health plan or retirement contribution — not just what the recommendation is. People need to see the reasoning.
• Autonomy means the system supports decision-making rather than replacing it. People should always have meaningful alternatives and real control.
• Data minimalism is about asking: "what is the minimum information we need to help this specific person make this specific decision well?".
• Transparency about incentives means being honest about how the platform makes money and how those structures do or don't affect recommendations.

‍

Q: SAVVI uses a prescriptive analytics engine rather than generative AI for its core guidance. Why?
Health plan decisions aren't a creativity problem. They're a math problem. When an employee is choosing between an HDHP and a traditional plan, the right answer is deterministic — given their healthcare usage, family structure, tax situation, and actual plan documents. What they need is a system that frames the problem correctly, runs the numbers correctly, and explains the reasoning clearly. A large language model optimizes for plausibility. Our guidance engine optimizes for accuracy. Those aren't the same thing.

Our core engine is built on prescriptive analytics — optimization-based, not generative. It produces consistent, auditable recommendations you can trace back to specific inputs and assumptions. It shows its work. It owns its uncertainty. And it gives the same answer to the same question every time, which is what you'd expect from any reliable analytical tool.

We do use generative AI — in the places where it belongs. Agentic multi-modal business process automation. Powering the conversational layer that makes guidance feel like a dialogue, not a report. Supporting internal workflows where domain experts are in the loop — people who understand the problem well enough to apply GenAI effectively and catch what it gets wrong.

The design principle: use generative AI where creativity and fluency are the job. Use a prescriptive engine where the job is to be right.

‍

Q: What does "shift-left ethics" mean in practice for your engineering team?

In security, "shift left" means finding and fixing vulnerabilities early in development. The same logic applies to ethics.

At SAVVI, every AI or machine learning technology goes through a mandatory pre-integration security and Ethical AI review before it touches our production environment. This isn't a checkbox exercise, it's a gate. We also run simulated population studies on our guidance algorithms across diverse employer populations to catch potential bias before it affects real people.

‍

Q: What's your response to the argument that ethical constraints slow down innovation?

If "moving fast" means skipping the work that builds user trust, you're moving fast in the wrong direction. It's like choosing to build a house faster by skipping the load bearing walls. A trust breach can set you back years, especially in financial services.

The platforms winning long-term in fintech are the ones that figured out how to scale trust alongside technology. That's the actual innovation challenge. Ethics sharpens focus and forces you to ask who you're really building for, which produces better products.

‍

Let’s find the right solution for you, visit savvifi.com or connect with us at sales@savvifi.com or 781 583 7011.