Process automation at the interface with engineering

PartSpace introduces an AI agent for technical procurement and cost engineering. The new software solution, PartsAgent, is based on technical data intelligence that consolidates CAD, PLM, ERP, supplier, and pricing data and makes it usable for operational decisions. On this basis, the AI agent can plan, execute and continuously optimize core procurement processes, from needs analysis and RFQs to order placement.

With this launch, PartSpace is introducing a solution that combines technical data intelligence with autonomous procurement. While many AI applications in procurement remain limited to text, dashboards or individual analytical steps, PartSpace relies on an end-to-end data analysis and execution layer for industrial procurement processes.

“Industrial companies have enormous amounts of technical and commercial data, but in practice, this information is usually disconnected and of little operational use,” says Sebastian Freund, COO of PartSpace. “This is exactly where PartSpace comes in: We make design data actionable, thereby creating the foundation for an AI agent that can independently execute key steps in technical procurement.”

From silos to executable processes

PartSpace employs a two-tier architecture for PartsAgent: At the data intelligence layer, CAD, PLM, ERP and supplier data are consolidated at the attribute level and converted into a structured data foundation for technical procurement. Building on this, the AI agent handles operational tasks in procurement and cost engineering, ranging from workflow planning to the bundling of RFQs, bid comparison, and order placement. To ensure end-to-end, consistent processes, PartsAgent is fully integrated into the customer’s existing system landscape using the open-source Model Context Protocol (MCP) standard. This enables the AI agent to connect to both internal and external data sources and actively execute relevant actions via MCP tools.

The solution identifies price anomalies at the part and supplier levels and supports strategic decisions, such as bundling similar parts. A typical use case: Open requirements for stainless steel turned parts with comparable geometry and material grades are automatically identified, technically bundled, and sent as an RFQ to suitable suppliers. Responses are recorded in a structured manner, quotes are made comparable, and savings potential becomes visible. In one example, the system bundles 12 parts with a total open quantity of 67,900 pieces and identifies a savings potential of 52,000 to 86,000 euros. PartSpace also combines technical and economic logic in the area of price intelligence. For example, the agent can specifically analyze conspicuous price trends and compare them with technical changes. In a sample prompt, the system examines parts with price increases of more than 25 percent over three years, identifies ten affected items, and assesses six of them as technically implausible. The optimization potential amounts to 180,000 to 320,000 euros.

Technical data as the foundation for autonomous procurement

PartSpace deliberately distinguishes itself from chatbots, GPT wrappers and pure reporting dashboards. At its core is a technical data intelligence platform on which an autonomous agent can operate. The solution is primarily aimed at purchasing managers, CPOs, technical buyers, and commodity managers, as well as those responsible for digital transformation and ERP/PLM landscapes. For PartSpace customers, 40 percent of the operational purchasing volume for drawing parts has already been fully automated. The goal is to automate the majority of processes so that professional buyers can focus on their core tasks.

“Many providers offer analytics, dashboards, or assistants. But the next step in technical procurement is a system that can act based on real engineering data,” says Freund. “It’s not the next user interface that determines the value of AI in procurement, but the quality of the underlying data.” For further automation in technical procurement, a consistent technical data foundation is considered a key prerequisite - from the design file through the quotation phase to the order. PartSpace brings together technical procurement and cost engineering on a shared data foundation to achieve this.