Compare models
Consider tasks, hardware and capabilities together.
Run AI on your own hardware, adapt open models and put company knowledge to work: on Tuesday, we show how this works in practice. With SAP and local AI, without data leaving the premises.
Themed afternoon · STARTPLATZ Düsseldorf · arrivals from 13:20 · sessions from 14:00
14:00–21:00 overall: themed afternoon and evening event
Consider tasks, hardware and capabilities together.
Make deliberate choices about sensitive documents and access.
Weigh the benefits and effort of your own system against external services.
From choosing a language model to running it securely: using NVIDIA DGX Spark, Yavuz Yıldız explains how local AI connects with SAP and makes company knowledge useful in-house.
View sessionProcess personnel data with AI without models or data leaving your organisation: hardware, model selection, HR tasks, access rights and data protection. Five theses based on his own measurements.
View sessionWhat happens when AI agents stop just answering questions and take over entire business processes? Real-world examples, digital sovereignty and a live demo of automated bookkeeping.
View sessionWhere Industrial AI fits into engineering, production and value creation, and what it delivers.
View sessionWhat an AI agent may decide on its own, who approves the rest, and how to prove it under the AI Act.
View sessionHow an AI platform runs on the customer’s own infrastructure and what that means for data sovereignty.
View sessionThe legal side of AI sovereignty: using open models, understanding dependencies and retaining control of your company’s knowledge.
View sessionAn open brain map, AI and a question: what can we experiment with? Using the fruit fly as an example, Hamid Hosseini explores the possibilities – and the ideas they could inspire for businesses.
View sessionExperience an open AI model running on a local computer. Compare how different models respond to the same task and discover what matters when assessing quality, speed and choosing a model.
View sessionWhich AI model runs on which device, and how fast? The session shows it live, from a microchip costing a few euros to a GPU server. Finally, you will see step by step how to start your first model on your own computer with Ollama and use it to answer an email.
View sessionFoundations – context – practice: what finance and controlling need to clarify before introducing AI, what matters during implementation and what decides success in practice.
View sessionA model alone does not get the job done. This practical session explores how to connect your own documents and useful tools, then check the answers against a concrete task.
View sessionWork through a live HR task using local AI on owned hardware, with pay transparency as the example. No installation required; bring your questions and use cases.
View sessionRight after his expert talk: how Industrial AI moves from pilot project into live operations.
View sessionWelcome at 13:40. Exact times for individual sessions will be added.
Go to the evening event ↓When should AI run on your own hardware?
Explore open models, local deployment, fine-tuning and further training. Understand how infrastructure, model adaptation and operational requirements relate to your use case, particularly when sensitive data is involved.
One application area: HR.
Expert talks and guided application sessions provide a basis for examining local deployment options. A supported project area offers time to work through the details.
For your projectEvaluate options for local AI and deepen your understanding of what operating it requires.
When is a local model the right choice? How do open models differ, and when are retrieval, fine-tuning or further training useful? The programme also addresses sensitive HR data, licences and the conditions for operating models.
Which tasks should use OpenAI or Anthropic, and which should run on your own AI so that critical company knowledge stays in your organisation? The discussion focuses on actual data flows and use cases.
Compare open models on the same task, work with your own documents, and explore deployment on a laptop, team server or European cloud. Explore the different approaches to find what suits your own requirements.
Specific sessions, contributors and technical requirements will follow.
Day passes – excl. VAT. The personal week pass brings all five days together.
Want to share your experience or run a hands-on session?
Help shape AI Week ↗