Christian Capdeville Joins Savant Labs: Delivering Trustworthy AI Automation

For the past fifteen years, I’ve watched knowledge workers drown in repetitive work.

I’ve been one of them. Early in my career, I spent more time hunting for data across systems, reformatting spreadsheets, and reconciling numbers than I did actually making decisions with any of it. The strategic work — the judgment calls, the pattern recognition, the “what should we do about this?” — always got squeezed into whatever time was left.

I wanted to help solve this problem, which is what led me to work at companies like Anaconda and Dataiku. I’ve been striving to close this gap for both technical and business teams. And I watched every wave of technology that promised to fix it — big data, machine learning, self-service BI — add new tools without truly eliminating the repetitive work.

The data is still scattered. The reports still get assembled by hand. The same manual wrangling still happens again next month.

But now, for the first time, I think that’s actually changing. There’s a path becoming clear — a way to actually close this gap using AI that’s grounded in something deterministic and trustworthy. Savant is building in this direction, and that’s a big part of why I wanted to be here. The implications of what’s becoming possible right now are worth paying attention to.

The Situation Right Now

Here’s what I see: a kind of two-class system is emerging.

On one side, you have teams in finance, tax, healthcare, or anywhere the data and outputs are sensitive and highly scrutinized. They’re largely still working the old way. Excel, manual processes, the same workflows they’ve been running for years. Not because they don’t want something better, but because they can’t afford the risk and hallucinations that they’ve seen from many modern AI tools. They can’t afford to take risks on technology that might get it wrong 5% of the time when the stakes are audit readiness or regulatory compliance.

On the other side, you have people actively trying to use AI to automate their workflows. But they’re running into the same wall: the non-determinism. Ask the same question twice, get different answers. They can’t fully trust it. So instead of eliminating the manual work, they’re just babysitting. Reviewing everything the AI produces, stitching outputs together, still spending hours getting to a result they can stand behind.

Neither group is getting what they actually need: automation they can trust to run without them, producing the same accurate result every time.

What’s Actually Working

What I’m seeing develop at Savant is different. It’s AI automation combined with an auditable, rule-based structure underneath. Users create deterministic workflows that enable consistent results every time, with embedded AI agents that understand intent and can handle messy, unstructured data. The rule-based structure gives you trust. The AI gives you speed and versatility. Together, they make it possible to actually automate the repetitive work, not just assist with it.

One of the things that convinced me Savant was the right place for me was seeing it work. Not a demo, but actual production workflows. Accounting teams that automated their month-end close. Tax teams that cut days off their timelines. People who’d been buried in spreadsheet chaos for years finally getting out from under it.

These aren’t pilot projects sitting in a sandbox. They’re running every month, handling real data, producing results people actually trust.

Why I’m Here

I’ve seen a lot of AI hype over the past few years. So have you, probably. A lot of it has made people skeptical, and rightfully so. Too many pilots that went nowhere. Too many promises that didn’t pan out.

I’m not here to add to that. I’m here because I think we’re finally at the point where knowledge workers can spend their time on the work that matters: the judgment calls, the decisions, the strategic thinking, instead of the mechanical work of hunting down data and moving it from one place to another.

If you’ve been promised that before and it didn’t deliver, I get it. But I think we’re at the precipice of something real, where people can finally stop wasting hours just arriving at the right picture of the information and actually spend their time on the work that matters.

That’s the future I am excited to help build, alongside my talented colleagues here at Savant.

– Christian Capdeville, Senior Director, Product Marketing