Startup Profile
Tentris

Introduce yourselves

Tell us in three sentences who you are

TentrisDB is the first scalable graph database purpose-built for GDPR-compliant and explainable machine learning. Our innovative data structure, powerful query engine and optimized ML algorithms enable up to 1000x faster queries and reduce data processing costs by up to 90%. Built on TentrisDB, our GraphRAG offers significant advantages over traditional RAG systems: It reduces token costs by up to 97%, minimizes hallucinations and delivers more accurate results for complex search queries. Tentris GmbH is based on more than seven years of research at the University of Paderborn.

Tentris is a spin-off from Paderborn University. From your personal experience, what are the advantages and challenges for a startup from university research?

Experience to date shows that our many years of research are a decisive advantage. It lends credibility to our technology and creates trust among potential customers, as the added value is recognizable. This facilitates market access and accelerates the validation of our solution.

An additional advantage is the deep understanding of our technology and its development over the years – knowledge that many first-time founders without an academic background have to build up painstakingly. At the same time, the transition from a research-driven approach to commercialization is a significant challenge. In research, the focus is often on edge cases and controlled environments with the aim of producing scientific publications. On the other hand, business aims to develop marketable products and sell them successfully.

Another important aspect is the step out of the familiar university environment into the dynamic, often uncertain world of a startup. However, those who manage to master these challenges and set the proper focus have set the course for a successful future.

Investors generally don’t have an AI or IT background. How do you explain your startup innovations to AI laypeople understandably?

Most investors we work with have a solid basic understanding of AI. This makes it easier for us to convey our technology’s depth and added value in detail. Our first pitch can usually be broken down into a simple one-liner:

“We are developing a tool – similar to ChatGPT – consisting of a powerful database and an intelligent AI assistant that enables companies to communicate precisely and hallucination-free with their fragmented data landscape in natural language.”

AI models are substantial energy guzzlers. Your technology promises to reduce energy consumption significantly. How do you do this, and who are your potential customers?

The secret behind Tentris can be divided into two central components: With TentrisDB, the magic lies in our unique data infrastructure and optimized query processing. Unlike conventional solutions, we do not rely on traditional table technology but on powerful tensor algebra.

TentrisAI, on the other hand, impresses with a drastic reduction in costs, which is achieved by minimizing the number of tokens required per query and using open source LLMs. The innovative knowledge graph storage format is at the heart of our solution, which seamlessly integrates structured and unstructured data. The key is that the better the data quality, the fewer tokens are required. How do we implement this so efficiently? That remains our secret. 😉

TentrisAI is aimed at companies with complex data landscapes that need to process demanding queries efficiently, expecting maximum precision and leaving no room for hallucinations.

What makes your startup unique or different from others on the market?

Our unique advantage lies in intelligent storage, efficient processing and seamless interaction with data. No other provider on the market combines semantics and cost-efficient scalability as convincingly as we do – two essential building blocks for future-oriented data management.

You received EXIST funding in the summer of 2024. What are the goals that you want to achieve with the funding?

The goal of EXIST funding is clear: to develop a functioning business model that will transfer our research into commercialization. In this way, we want to contribute positively to the importance of spin-offs from research and show that there are great programs in Germany that support start-ups on this path.