Explore our Services

AI Exploration

Stay AI-curious. Navigate your journey.

Ready to start your AI journey, but overwhelmed by the sheer number of options?

We help you navigate the AI landscape and identify all the necessary steps to bring your project to success. Together with you, we assess which tools are the right fit for your needs and connect you with relevant stakeholders who can help you implement your ideas.

This service is designed for anyone who is interested in AI and looking for an initial introduction to the field. In our consultation, we focus on the most important concepts, such as traditional machine learning (ML), large language models (LLMs), retrieval-augmented generation (RAG), WebUI/Blablador, and AI agents.

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

Data Consulting

Create AI-ready data for AI-ready organisations.

Do you have a large collection of heterogeneous data and wonder whether it can be used for AI applications?

Join one of our consulting sessions and let us find out together. We assess your data readiness for AI and help you identify its potential. Our experts support you in cataloguing and preparing your existing data, as well as setting up on-premises data pipelines and data governance frameworks.

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

Inference Consulting

Choose the right inference option for optimal performance of your model.

Not sure which inference solution is the right fit for your project?

You've come to the right place. We help you identify the most suitable inference approach for your AI application and advise you on how to optimise your model's inference performance. Our team offers access to and support for a range of inference services. With JARVIS, the AI Factory JAIF will design an experimental AI inference module for the JUPITER AI Factory, based on cloud technologies and complementing the JUPITER training system. In addition, we provide an evaluation server for large language models called Blablador. We are happy to introduce you to Blablador. Learn more about its web interface, API, default models, and customisation options. We are here to help you get started.

Explore Blablador

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

Knowledge System Consultancy

Be the architect of your organisation’s knowledge framework.

You are thinking about developing a reliable AI assistant grounded in your own data?

Our AI consulting service, with a focus on knowledge systems, may be of interest to you. We advise you on flexible knowledge architectures, such as semantic search, simple retrieval-augmented generation (Simple RAG), as well as combinations of knowledge graphs and retrieval-augmented generation (KG-RAG), tailored to your specific use case.

The knowledge systems we develop are transparent, trustworthy, and fully compliant with European data protection standards. They provide grounded answers with traceable source references, allowing users to verify the information while significantly reducing hallucinations. Time- and cost-effectiveness are important to us. Our knowledge architectures typically do not require extensive model fine-tuning or retraining.

How does our consulting process work in detail?
Together with you, we create strategies for implementing guardrails and evaluate the AI agent best suited to your use case. We help you assess your data readiness and optimise your knowledge base. Our experienced team draws on a wide range of best practices for agentic workflows and tool integration, supporting you with model selection, prompt engineering, and the deployment and fine-tuning of Small Language Models (SLMs).

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

AI-Model Training Consulting

Our supercomputing power for your AI model.

How can you optimise the performance of your AI model?

Train your model on Europe's strongest supercomputer, JUPITER. By leveraging exceptional computing power, large memory capacity, and high-speed data processing, supercomputers significantly advance model training and enable the development of larger, more sophisticated AI models. JAIF combines expert guidance with access to state-of-the-art high-performance computing (HPC) infrastructure to help you train AI models efficiently. We advise you on scaling AI workloads, optimising data loading and node utilisation, and ensuring compliance with AI safety standards. From training preparation to final fine-tuning, we guide you through every stage of the training process.

We work with the following tools:

• PyTorch: An open-source framework for deep learning
• TensorFlow: An end-to-end platform for machine learning to create ML models
• Nsight: A collection of GPU profiling tool set provided by Nvidia
• Slurm: Slurm is an open source, fault-tolerant, and highly scalable cluster management and job scheduling system for large and small Linux clusters
• NCCL: The NVIDIA Collective Communication Library (NCCL) implements multi-GPU and multi-node communication primitives optimised for NVIDIA GPUs and networking.
• CUDA: The NVIDIA CUDA Toolkit provides a development environment for creating high-performance, GPU-accelerated applications.

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

AI Agent Consultancy

Streamline your workflows and speed up your decision-making

Are you considering implementing an AI agent in your organisation to streamline your workflows?

In comparison to chatbots, which simply respond to user prompts, AI agents act more autonomously. Agentic AI is an artificial intelligence system that can imitate human decision-making and can be used to improve or take over entire workflows. Our experts will advise you on how to build and use powerful AI agents in scalable and trustworthy environments.

Selected opportunities:

• Agents via JupyterLab and JSC Cloud
• Hosting n8n: agent.blablador.fz-juelich.de
• Consulting on agentic use of LLM’s, using Blablador
• Coding Agents (Hermes, OpenCode)

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.

Compute Consulting

Find your match and get the right compute resource for your project.

Do you work for a company, a SME, or a university and want to test a new AI application on one of Europe's High Performance Computers?

We are happy to help you identify the best way to access the compute resources you need and review your proposals for computing time. We work closely with the EuroHPC Joint Undertaking (EuroHPC JU), which offers a range of access schemes through its AI factory access calls. In addition, we collaborate with a broad network of national public and commercial providers that offer access models tailored to your needs.

In particular, our partner hessian.AI provides access to a unique high-performance computing (HPC) infrastructure that is made available to companies and research institutions through dedicated calls.

You can find the current EuroHPC JU calls here

Learn more about the available hessian.AI calls here

Interested? Get in touch with us to schedule an initial consultation with the JAIF expert team.