Interview with PartSpace CEO Robert Hilmer

"Our customers benefit from tailored purchasing decisions, transparent processes, and significant savings."

Robert Hilmer is the CEO of PartSpace. He founded the company in 2019 together with Sebastian Freund (COO) and Michael Neuhauser (CTO). In an interview, he answers questions about the company, the AI software solution, and other topics.

What was the founding idea behind your company?

Robert Hilmer: Before founding the company in 2019, the founders had been working in technical purchasing in the field of special machine construction and mechanical engineering for years. That's why we know that purchasing components and assemblies that have to be specially manufactured based on a design file is highly complex. This is due, on the one hand, to the lack of transparency and fragmentation of the market and, on the other hand, to the diversity of components and the fact that buyers' knowledge is often not documented centrally anywhere.

This gave rise to the idea of developing a specialized purchasing platform for manufacturing parts and assemblies. Our goal was to identify the optimal procurement solution for every drawing part and every assembly and to efficiently connect companies from industry and medium-sized enterprises with the right suppliers.

How did PartSpace develop after it was founded? 

Robert Hilmer: Early on, we received inquiries from medium-sized companies regarding similarity searches and the “understanding” of 2D and 3D data, which we were able to serve and took as an opportunity to further expand our AI expertise. The potential of our technology was also recognized by our first major customer, a global agricultural machinery manufacturer. The software rollout at the agricultural machinery expert's first plant delivered impressive results, leading to a global rollout. For us, this meant the consistent further development of the technology and, as a result, further successful implementations of our AI platform at well-known companies such as Siemens and Krones.

What is your vision?

Robert Hilmer: Our vision is to enable companies to use engineering data as a central basis for decision-making and to have the procurement of designed components and assemblies carried out entirely AI-controlled – from design to ordering.

How would you describe PartSpace’s AI solution?

Robert Hilmer: Our AI software, PartSpace AI, understands technical drawings, analyzes design data fully automatically, and uncovers previously hidden savings potential in the procurement of drawing-related components. The AI model used for this was developed specifically for technical purchasing. While classic AI systems primarily rely on text or ERP data, our AI has been specifically trained to recognize relevant features in CAD drawings and 3D models – from dimensions and tolerances to materials and post-processing methods. The AI automatically analyzes, structures, and links this information with historical purchasing data and external market information. It recognizes how a component can be manufactured and what purchase price can be achieved from which suppliers. The result is a digital assistant that not only processes engineering data automatically, but also evaluates it strategically and makes it usable. 

In which areas is the solution used?

Robert Hilmer: The structured data resulting from PartSpace AI opens up a wide range of applications: from strategic purchasing, cost engineering, and supplier benchmarking to standardization and quality assurance. Our software not only can calculate realistic target prices, but also identify suitable suppliers or detect duplicate components through similarity analyses. This allows variants to be bundled and volume effects to be exploited in a targeted manner.

In addition, we are working with our first customers to handle the entire purchasing process via our AI – from targeted supplier inquiries and automatic renegotiations to the ordering process. 

 What makes PartSpace AI so special?

Robert Hilmer: What makes it special is that the PartSpace AI model was trained using real drawings, prices, supplier data, and processes from actual, individual customer projects, rather than synthetic or publicly available data sets. This historically grown database ensures particularly robust analyses and reliable results. Each company receives an individually trained AI model based exclusively on its own data.

The individually trained AI model is based on customer data – how is data security guaranteed? 

Robert Hilmer: Data security is our top priority. Our systems run in protected cloud environments that meet the highest security standards. Companies retain full control over their sensitive information at all times while enjoying the benefits of intelligent automation.

Can you outline an example of how it is used?

Robert Hilmer: For example, a machine manufacturer used our product to perform a drawing and cost analysis of product groups using our AI:

• Initial situation and challenge: The customer had over 10,000 drawing-related components, which were divided into different product groups of around 5,000 parts each for better structuring. PartSpace AI was used to implement an efficient similarity and cost analysis for this large quantity of drawing parts.
  

• Implementation: First, the design data for an initial product group was provided and around 2,500 components were implemented. PartSpace AI automatically extracted the cost-driving properties from the technical drawings and linked them to geometric features from the 3D models.
  

• Results: The first results were available just a few weeks after the software was introduced: The AI accurately predicted the prices of all components, identified price deviations, and suggested the optimal supplier for each component from the entire supplier base. After extensive data analysis of the first two product groups with a purchasing volume in the double-digit millions, a savings potential of around 12 percent was identified within a few months. 

The conclusion: PartSpace AI contributed significantly to making the similarity and cost analysis of drawing-based manufacturing parts quick and easy and to achieving price transparency across all components. The return on investment was achieved within a few weeks, and further cost-saving potential was subsequently realized by analyzing other product groups.

AI requires enormous expertise and proven experts. How many employees at PartSpace are responsible for AI technology?  

Robert Hilmer: We have a strong team with comprehensive AI expertise for the industry. More than 25 specialists in the fields of artificial intelligence and engineering - including four PhD experts who have published relevant scientific publications as part of their dissertations - contribute their in-depth knowledge and passion to set new standards in the application of AI for technical purchasing.

How do you see the company and its solution positioned? 

Robert Hilmer: With this solution, PartSpace is positioning itself not as a traditional tool provider, but as a strategic partner for industry, combining technical expertise, trustworthy data architecture, and real-world experience. While many companies are still experimenting, PartSpace AI is already in productive use, has proven itself on the market, and delivers measurable results. This shows the potential that lies in combining technical expertise and AI. PartSpace is one of the pioneers in the use of AI in engineering. Its clear goal is to make technical purchasing smarter, safer, and more sustainable.

Our customers benefit from customized purchasing decisions, transparent processes, and significant savings through optimized procurement strategies. 

 What are the next steps in the development of PartSpace?

Robert Hilmer: The next goals are to further develop our AI, roll out internationally, and accelerate the company's growth. To this end, we are working with Armira Growth and Bayern Kapital. With the support of our investors, we will drive forward scaling and enable our customers to make better purchasing decisions.