Talk
On-Premise AI with Local Models
OpenAI and Anthropic dominate the AI landscape. Unfortunately, all data sent to their models is processed by these third-party providers. That makes it quite a challenge when it comes to internal data, customer data or anything else where GDPR applies.
Local LLMs, which you can run inside your organization, seem to be the obvious solution. In this talk, we discuss how to set them up, how to embed them into your applications, what costs to expect, and what level of output quality you can expect. We will also present you with real-world examples from different fields like manufacturing, business services or even software development.
You’ll leave this talk with inspiration for your own projects and an honest understanding of what it takes to use local models in practice.
About Rainer Hahnekamp
Rainer Hahnekamp is a Google Developer Expert (GDE) and AI Architect at Soverius AI, with over 15 years of experience building mission-critical enterprise systems.
At Soverius, he works with enterprises to run AI on their own infrastructure and data, specializing in local LLMs, RAG, harness engineering, and agentic development.
A particular focus is AI coding assistance - coding assistants that run on a company's own models, so proprietary source code never leaves its environment.
He has an academic foundation in Unsupervised Machine Learning from the University of Vienna.