The AI Services is a project at Hasso Plattner Institute (HPI) funded by the Federal Ministry of Education and Research (BMBF). Our goal is to lower the barriers to using AI in business and society.
We support start-ups, companies, public institutions, students, and researchers in developing and using AI applications. To this end, we provide the necessary resources, training, and advice at no cost.
Our services at a glance:
Simplifying AI Transfer
AI Services at HPI is one of four AI service centers in Germany that are funded by the BMBF.
With our AI Services, we address those, in particular, who are interested in machine learning (ML) and who have already taken the first step and are now faced with the question of how to move from an initial trial of ML to productive use.
Prof. Holger Karl, Scientific Director of the AI Services and the Internet Technologies and Softwarization research group at HPI
Fields of Consultation
We offer individual consultations in which we address your particular needs. You are welcome to book these meetings with us using the following link. Alternatively, you can also send us an e-mail.
Contact us
With our AI consulting, we support organizations free of charge in implementing their own AI applications. In particular, we can advise you on using our computing infrastructure and provide methodological advice on selecting AI model architectures, AI software frameworks, and the implementation of AI applications, considering the latest scientific research results. AI consulting consists of several phases. First, we want to understand your use case; this means to identify (1) and specify (2) it. Then, we develop (3) a joint solution, which we finalize through documentation (4) in the last phase.
Pilot projects are a special case of AI consulting. For us, l pilot projects are the prototypical implementations of ideas that organizations bring to us and that we implement together. They are prototypical because we cannot provide a productive operation of applications. The prototype is not commercially viable but can be further developed and thus commercialized.
For organizations or individuals who do not have a specific AI use case they would like to implement with us, but who nevertheless have a connection to AI, we offer cooperation in the form of jointly designed networking events or educational offers such as workshops, lectures or Massive Open Online Courses.
We offer to work with start-ups during their development process and support them with networking and AI consulting, pilot projects, and collaborations.
Education
We offer a range of face-to-face, online, and hybrid training courses. We are constantly developing our training courses. We are always open to suggestions – please send us an e-mail.
Contact us
Our formats:
Our workshops are aimed at medium-sized groups of 5 to 25 people from various organizations.
Our AI Maker Sessions are a series of comprehensive workshops for anyone who wants to deepen their knowledge and skills in artificial intelligence. They take place every Tuesday and Thursday. The workshop series provides a comprehensive insight into theoretical concepts and practical applications of various AI-related topics, such as training and evaluating machine learning models, neural network architectures, and MLOps. In addition, participants get prepared for our planned certification program for AI experts.
Our weekly paper reading sessions are aimed at AI users, data scientists, and software engineers. These workshops cover the latest developments in automatic natural language processing, computer vision, and related fields.
The Mittelstand Digital Center Berlin at HPI offers a practice-oriented and interactive introduction to artificial intelligence. In the workshops, basic concepts and applications are taught using real-world examples.
With our AI Service Center channel on the openHPI platform, we offer a selection of online courses on various AI topics. Upon completing the courses, you can receive certificates that confirm your learning success and attest to your AI expertise.
We prepare an AI certificate that you can use to certify your knowledge of AI topics.
We invite you to recurring Guest Talks on-site at HPI. Under the heading KISZ-Guest Talks on HPI tele-TASK, you will find in-depth talks on current AI topics from research, development, and business. In the future, we will also offer our own talks, which are aimed at a broad audience from society and business without requiring much prior knowledge and intended to fundamentally classify AI and explain its opportunities and risks.
Can't find what you're looking for?
Couldn't find the right educational program for you? Then visit our partners: the HPI School of Design Thinking, the HPI School of Entrepreneurship, the Mittelstand Digital Center Berlin and the HPI Research Schools – or subscribe to our newsletter. We will inform you about current workshops, job offers, and other events every month.
Become part of the AI Maker Community
Computing resources
The AI Services at HPI provide a heterogeneous AI infrastructure. This infrastructure is used to research the requirements and needs of different interest groups, such as academic institutions, start-ups, or small and medium-sized enterprises. With this infrastructure it is also possible to compare the costs and performance of different types of hardware, investigate edge applications, and provide flexible solutions for the development, optimization, and implementation of complex AI models in production environments.
The most powerful hardware is 8 NVIDIA Basepods with 8 H100 GPUs each (with 80 GB VRAM each), which can be used for training AI models. The H100 pods are connected to a fast memory system that enables direct writing from GPU memory to persistent memory. The H100 pods communicate with each other via 400 Gb/s Infiniband or 200 Gb/s Ethernet. For inference, 5 NVIDIA pods with 8 A30 each are available, which communicate with 25 Gb/s Infiniband or 40 Gb/s Ethernet. In addition, we offer access to an NVIDIA Jetson AGX module, an ARM server, a medium-sized server, virtualization servers, and storage modules.
To use the resources of the AI Services, please send us a brief project description. If your project is approved, we will require you to sign a user agreement, which is a crucial step in accepting our terms of use.
We will also be happy to help you select suitable resources and inform you about the advantages and disadvantages of cloud, on-premise, and hybrid solutions.
Request via e-mail
Commitment of resources
Signing of the user ge contract and creation of the conditions of use
Access to computing resources
Documentation of the results and findings
Last change: 17/09/2024, Mareike-Vic Schreiber