Special Edition: Security in LLMs and Gen AI Apps @ Celonis

✨ Special Edition: Celonis treats to coffee and breakfast ✨

Generative AI is powerful — and risky. Prompt injection, data leakage, jailbreaking, and malicious use of embeddings are just a few of the threats we face when building AI-driven systems.

In this session, we’ll swirl around:

• Real-world security risks in LLM-based applications
• Common attack vectors and how they work
• Practical strategies for mitigation and defense
• Open questions where the industry still lacks answers

This time, Celonis kindly provides location, coffee, and breakfast.

🕙 Meeting time: 07:50
📍 Meeting point: Theresienstraße 20, in front of Celonis doors.
🚪 RSVP closes on the 10th of September at 13:00.

Note: The Organizer will ask for your full name and email upon RSVP-ing in order to arrange access to the Celonis office.

No slides, no sales pitches. Just an honest conversation about what it takes to build secure, resilient AI systems.

Time: 08:00 (yes, early — but we start before the day steals your focus). Small group, every person matters. Can’t make it? Please update your RSVP.

Language: English 🇬🇧

Format: One topic, a bunch of friendly people, open discussion. There’s no speaker. There might be prepared questions for the round, but the discussion may derail — and that’s fine. Whatever happens is the only thing that could have happened.

Entry level: None. Join if you’ve never built a reactive system but want to learn, or if you’ve been building them for years and want to defend or challenge best practices.

Note: The Event is privately organized. Not an official Celonis event.