Etnology #6: What does AI know? And who let it loose?
Etnetera, Jankovcova 49, 170 00 Prague 7 - Holešovice
.webp)
Pavel Pola
Martin Gyurák
Filip Beneš
Tereza Fukátková
David Bečvařík
Jiří Štěpán
Ondřej Franek
Vladimír Beran
David Hujer
Antonín Šimek
Jan Zubíček
.webp)
Radek Tůma
.webp)
Data you don't trust. Agents you don't know about.
Almost everyone is trying AI now. Only a fraction reaches production. Two things decide it, and they're the ones nobody looks at. First, data quality. An agent works only with what it finds in the data. Second, people. Employees use AI anyway. The company often just doesn't know.
Data you can't vouch for
Customer data sits in three systems in three versions. The definition of an "active client" changes depending on who you ask. No licence fixes that. With a chatbot it meant a wrong answer. With an agent that has to act, it means the work itself comes out wrong. A wrong order, wrong pricing, the wrong answer to a client.
Agents you don't know are live
An employee builds an agent in an afternoon, gives it access to email or the CRM, and sets it running. That's a process no one approved and no one is watching. When the person who built it leaves, no one knows how to switch it off. This happens even at companies where AI is officially banned.
What you'll take away
Both are manageable. On two panels, people from companies that are actually dealing with this talk straight about what works and what doesn't. In the demo zone, solutions run live in production. You leave with a concrete sense of how to build an official path people will follow on their own.
The agent is just the tip of the iceberg
At Etnology #5 we covered agents and the platform.
At #6 we go deeper, with the experience that without ready data and adoption under control, the platform alone isn't enough. The agenda is built on a mix of keynote talks, panel discussions with AI experts from client companies, and a live AI demo zone.
Agenda
14:30 Doors open, tech demos
Welcome and AI landscape overview – Jiří Štěpán
The frame for the whole afternoon. How bad data and the AI people run on their own are connected.
Keynote: Data quality and data readiness – Pavel Pola
A model is only as good as the data it stands on. Data readiness can't be skipped or bought; catching up later is hard and expensive.
Panel discussion: How to improve data readiness
Data readiness sits between IT, business, and compliance. This panel is about who actually owns it, who pays for it, and what that means for AI adoption.
16:00 Coffee break
Keynote: Shadow AI – Vladimír Beran
Employees build their own agents and point them at company data. It's a security, regulatory, and operational risk all at once. We'll cover how to handle shadow AI without smothering adoption.
Panel discussion: How to speed up AI adoption
How to build the official path quickly and safely. This is where data and shadow AI meet in practice, with examples of current best practices.
17:40-20:00 Refreshments, tech demos and networking
Each demo comes with a short elevator pitch. Real solutions from production. Process automation, leaner workflows, better UX, straight from our AI experts. Ask questions, join the discussion, take away the know-how.
- Etnetera Agentic: complaints handling agent
- On-device AI with Apple Foundation Models in the Guest Connect app
- GEO with Profound: who shows up in the answer instead of you
- Škoda Customer Vibe CX platform
- QA Control Center: a new approach to testing in the age of AI
Speakers
.webp)
.webp)
.webp)
.webp)
