To close the day, we hosted a panel featuring Rob Straathof (Liberis), Camille Rougie (Plural.AI), and Sufyaan Kazi (Google), moderated by Bella Renney-Thwaites. The discussion was honest, sharp, and full of practical wisdom for anyone building with AI today.
We’re already in the Agentic Era
AI is no longer a back-office tool; it’s operating at the core of real business processes. Liberis has kept headcount flat at ~240 people since 2021, but grown revenue nearly 10x. That’s not just automation, that’s operational leverage driven by agentic systems.
But maturity is uneven
Only 11% of fintech leaders feel confident in their responsible AI frameworks. Regulation, risk, and data governance remain major blockers, especially in Europe. The winners will be the ones who build transparency and oversight into their AI from day one.
Not every use case needs generative AI
Camille Rougie reminded the room that not all problems need large language models. In many workflows, a simple deterministic system is more reliable and more secure. Knowing when not to use GenAI is just as important as knowing where it can shine.
The real value lies in memory and explainability
Agentic AI isn’t about speed alone — it’s about systems that learn from past interactions and apply that memory to future decisions. Both Google and Liberis are focused on building explainable systems that offer transparency, auditability, and accountability.
Co-pilots are coming for SMBs
Rob laid out a bold vision: what if every small business had its own AI CFO, CMO, and COO — embedded inside a partner platform? Systems that help merchants negotiate funding terms, optimise costs, and manage marketing budgets. Not someday. Soon.