We founded scitlab in 2020 with the vision of making AI accessible for European SMEs – simply and securely. We want to share more about the challenges and opportunities in the SME sector, our technical approaches, and everything we encounter building an AI startup.
5 / 5
How we built a flexible open-source stack for language models with LiteLLM, OpenWebUI, local models, and automated data anonymization — without vendor lock-in and with full control over our data.
Read moredoccape combines rule-based methods with AI to reliably detect sensitive data in texts, images, and embedded content. Consistent pseudonyms preserve context — ideal for analytics and AI applications. The solution is customizable and can be deployed locally or on EU servers — for maximum control and privacy.
Read moreAnonymization or pseudonymization — what's the right approach in the age of AI? In this post we explain what sets these two methods apart, what legal frameworks apply, and how you can still conduct valuable data analyses despite privacy requirements. With examples, use cases, and clear recommendations.
Read moreLanguage models have become indispensable in many workplaces. Yet companies still face open questions and hurdles — especially when sensitive data is involved. An in-house company chatbot is a real alternative to off-the-shelf solutions. OpenWebUI offers an intuitive interface ideal for integrating various language models, and brings many advantages with it. In this article we explain why we use OpenWebUI and why we integrate doccape into the processing pipeline.
Read moreIn today's data-driven world, AI workflows are growing in importance. At the same time, requirements for data privacy and secure data processing are rising — especially when sensitive information is involved. With doccape and our new n8n module, we offer a simple and secure way to seamlessly integrate anonymization and pseudonymization into your automated workflows.
Read more
