AI offers significant productivity gains through tools like co-pilots but also raises concerns about job displacement and code quality.
Developers must cut through the hype by focusing on foundational principles and best practices to select valuable tools from a rapidly growing ecosystem.
Zalando applies AI across its business, from customer-facing conversational AI to back-office forecasting, demonstrating that maturity is measured by time-to-value.
Building robust AI products requires a systems-thinking approach where agentic workflows and components like RAG are integrated, not just focusing on the core model.
The journey from a working prototype to a production AI system involves a difficult cycle of prompt engineering, MLOps, data labeling, and continuous monitoring.
The Machine Learning Development Lifecycle (MLDLC) adapts proven SDLC principles to AI, integrating data science and DevOps to manage risk and standardize tooling.
AI systems require specialized monitoring for statistical performance and data drift, alongside strong data governance and adherence to emerging compliance regulations.
Practitioners must recognize that AI is a specialized tool and should avoid forcing an AI solution when a simpler, non-AI approach is more effective.
msg
Ismaning, Germany
45:09Gunnar Grosch
43:05Kenneth Auchenberg & Christian Heilmann
23:55Zan Markan
47:35Laurie Voss, Rey Bango, Hannah Foxwell, Rizel Scarlett & Thomas Steiner
1:09:49Simi Olabisi
30:49Marin Niehues
58:06Anna Fritsch-Weninger
41:06Lars Gentsch
Jobs that call for the skills explored in this talk.
AzuraIQ Technologies GmbH
Graben-Neudorf, Germany
