Guillaume Beauverd

Beyond RAG

At the Panoramai AI Summit, Guillaume Beauverd, representing Candice following recent €1 million funding, presented a revolutionary framework for enterprise AI adoption. Drawing on 20 years of digital transformation experience across banks, FMCG, and healthcare companies, he advocated for organic, user-driven AI deployment over traditional top-down IT implementations.

Digital Transformation Evolution Beauverd positioned current AI adoption within historical transformation context: "So for the past 20 years, me and my colleagues and... We've been working on digitizing large companies, from banks to FMCG players, healthcare. I mean, we basically bring together a wealth of information about what it actually means to transform an organization and to actually implement tech."

Traditional Implementation Failures His analysis of past approaches revealed systemic problems: "So what we've gone through over the years from waterfall projects with large budgets, with gigantic teams, plans, and then rollouts that take time and don't necessarily succeed, because most of the cases the IT was in charge. Not to say that the IT is not capable, but the size of what was in their hands was very big and very often we felt there was a gap between what the business was actually looking for and what the IT had to deliver."

AI as Organizational Reshuffler Beauverd's central thesis positioned AI as a fundamental change catalyst: "And I feel that now AI is reshuffling the cart. It's changing the way we should approach AI deployments, technology deployment within enterprises. And for that matter, my point is, for the first time, it may be possible for business users to choose and pick what tools augments them, enable them."

Organic Advocacy Model His proposed adoption framework emphasized user-driven transformation: "And as they do so, they'll become the advocates of the value that they bring, and so they can very easily then convince others. By convincing others, that's hopefully gonna create some type of platform, if you wish, intelligence platform, that instead of, as we would have done a couple of years ago, define a process end-to-end and try to roll it out down the throats of the users and the business users, it's actually gonna organically, kind of Darwinistically transform the organization."

Power User Strategy His implementation approach focused on strategic user selection: "There's no real need now for a POC, for a pilot. It's just to let the AI do the one thing that it can do very well. And in particular, try and put it in the hands of the power user that's gonna advocate for it and basically change the organization with it."

Trust and Precision Requirements Beauverd acknowledged Swiss expectations for AI reliability: "But all of that needs to be trustable. And our test with RAG and other technologies show that it's very difficult to get to a confidence level that we Swiss people would expect, so an exhaustive result, a precise result, and this is the mission that we've gone working on with Candice."

Grunt Work Elimination Mission His company's ultimate vision addressed fundamental work transformation: "It's basically defining what precision should look like, what exhaustivity should look like, and build a tool that you can trust, that you can explain, and fire them at grunt work, the work that humanity should have never started with. So this is our mission, is to basically get the world rid of every grunt work out there, and enabling people make better decisions."

AI Performance Reality Check Beauverd provided honest assessment of current AI capabilities: "We are using a lot of AI. And honestly, to the discussion I had earlier with someone, There are things that AI does extremely well, but at the moment there are things that AI does less well than it did a couple of months ago. And so we are revising some of the aspects of what we had automated with AI in order to keep the quality or our expectations at least where they are."

Team Scale Achievement His company's accomplishments demonstrated AI-enabled productivity: "It's a good point. We are We're a team of five, and we've built something that is utterly complex."

Key Takeaway: Beauverd demonstrated how successful AI transformation requires empowering business users to drive organic adoption through advocacy rather than relying on traditional IT-mandated implementations, while maintaining Swiss standards for precision and trustworthiness.

Founder of Candice AI and seasoned management consultant with over 25 years of experience solving complex strategic challenges across various industries. He will introduce Candice AI, a document intelligence platform that automates the extraction and structuring of information from unstructured data—such as contracts, compliance reports, and supplier documentation. He will share how Candice enables organizations to shift from manual checks to automated insights at scale, freeing teams to focus on strategic work. Guillaume will highlight how Candice’s exhaustive extraction and traceability deliver the precision and reliability needed for trusted AI deployment in critical business processes. A serial entrepreneur, Guillaume has co-founded several successful ventures, including DualNetworks (IoT sensors and analytics), OPCIS Services (risk and compliance software for banks), and Trace Technologies. As former Managing Partner at mas advisory, he led teams supporting management, boards, and investors in areas including strategy, operational excellence, digitization, and risk management. Based in Geneva, Guillaume has been recognized as one of the top 100 Swiss personalities by l'Hebdo magazine and named among the 100 most innovative Swiss under 40 by Bilanz.