Direct procurement’s next data platform starts with a drawing

Direct-material procurement has a data problem that conventional spend analysis was never designed to solve.

An invoice can tell a team what it paid, which supplier received the order and how the transaction was classified. It cannot reliably explain whether two differently named components are technically equivalent, whether a specification is creating unnecessary cost or why plants pay different prices for materially similar parts.

That gap helps explain the latest investment in CADDi. The manufacturing AI company has raised a reported $114 million Series D at a $1.2 billion valuation, with investors including Moore Strategic Ventures, Coreline Ventures, Salesforce Ventures and Toyota’s Woven Capital.

CADDi’s platform is built around data normally fragmented across engineering and manufacturing systems: drawings, specifications, bills of materials, product lifecycle systems, ERP records and internal files.

The procurement relevance is straightforward. Direct-material opportunities are often hidden behind inconsistent part numbers, descriptions and technical language. A category manager may see thousands of separate line items even where engineering would recognise families of similar components.

Connecting the commercial and technical records can expose price variation, duplicate sourcing, reusable designs and specifications that make a component unnecessarily expensive to produce. It can also give engineering and procurement a shared evidence base before either side approaches a supplier.

Most spend platforms create value by cleaning supplier names and classifying transactions. That works well when the analytical question is how much the organisation spends with a supplier or category.

It is weaker when the question is whether two machined parts should cost the same, whether a previous design could be reused or whether a tolerance is restricting the available supplier market.

Those decisions require geometry, materials, tolerances, process steps and manufacturing history. The important market signal in CADDi’s funding is therefore not simply that another AI company has raised money. It is that investors see a valuable independent data layer between engineering, operations and procurement.

That creates a strategic architecture question for manufacturers: should technical-commercial intelligence sit inside the ERP, inside the source-to-pay suite, or in a separate manufacturing platform capable of working across both?

The financing does not prove CADDi’s customer outcomes, and its product claims still need validating in live deployments. But the underlying problem is real.

Direct procurement cannot reach its full potential while the drawing, the purchase history and the supplier decision remain in different systems.

Sources: Fortune · The Wall Street Journal · CADDi