Glad you're here! We've had a busy few weeks. PwC made a big move, we've got a new Jubel article out, and there's a stat in here that might make you rethink how your team works with AI. Grab a coffee and let's get into it. ☕
NEWS
73% of clients now expect real-time visibility into project status
Kantata's industry report found that nearly three quarters of clients demand real-time project visibility, while 89% of services leaders still spend significant time manually verifying AI outputs before sharing them.
PwC launches PwC One
PwC introduced PwC One, a platform combining its proprietary methodologies with AI models from Microsoft, Google Cloud, AWS, OpenAI, and Anthropic, letting clients interact directly with autonomous agents for tax analysis, deal diligence, and sustainability assurance.
What can we learn from this?
PwC just made 364,000 employees share the room with an AI agent. Not because AI is interesting, but because having juniors rebuild the same analysis for the fifth time doesn't scale. PwC One stores institutional knowledge in the platform instead of in people's heads. Smart move. But it also exposes the real gap: most firms have no structured way to retain what they learn. New engagement? Start from zero. New handoff? Context gone.
The Big Four can build custom AI platforms. Mid-market firms can't. So the firms that win won't be the ones with the best AI. They'll be the ones whose operations were already ready for it.
BEHIND THE SCENES
We are releasing the first feature of our new product this week 👀
Something's coming. This week, we're dropping the first feature of something we've been building quietly. Want in before everyone else?
👉 Don't miss out, join the waitlist
Bye bye Nora. Happy 30th Mathias.
Two milestones. We said goodbye to Nora after a great internship (she left us better than she found us, which is exactly how it should work). And we celebrated Mathias' birthday with the team. Cake was involved. Strong opinions about cake were also involved.
New Jubel article is out
The jam problem: Why professionals stick with bad tools.
73% of audit firms call themselves digitally transformed. Only 19% are happy with their own client-facing tools. The gap isn't about technology. It's about the behavioral patterns that keep firms locked into what they've always done. Read the article, second in the series.
👉 Dutch version 👉 English version
INSIGHTS
"89% of services leaders still spend significant time manually verifying AI outputs before sharing them with clients."
(Insight from Kantata | State of the Professional Services Industry Report)
🔍 The Systems View
Toto De Brant, Co-founder Alkmist
If you're spending hours double-checking your AI, you haven't built verification into the workflow. The answer isn't more review. It's structured outputs with built-in validation.
💬 Behavioral Perspective
Mathias Celis, Co-founder Alkmist | PhD Psychology & Business Economics
This is a trust gap, not a tech gap. People verify because they're afraid of looking stupid in front of a client. Fix the fear first, automate the checks second.
We're looking for an AI Engineer
*Master's in CS/AI/ML required.
TIPS
If a client asks you to resend a file, your process already failed. Keep every document exchange in a single, shared location with activity tracking. The question "did you receive it?" should answer itself.
Come Find Us at Accountex | 13-14 May | London
The UK's biggest accounting event is almost here, and we'll be on the show floor ready to chat, demo, and maybe even share a few secrets.
See you for the next one!
