Procuros Company Spotlight: Building Autonomous B2B Trade

One company's purchase order becomes another company's sales order. Then it becomes a warehouse instruction, a shipping notice, an invoice, a payment obligation, and, if one product code or quantity changes along the way, somebody's afternoon.

Procuros is building the connectivity layer meant to keep that chain intact. Founded in Hamburg in 2021 by Patrick Thelen, CEO, and Benjamin Wulff, CTO, the company connects suppliers, retailers, logistics providers, and ERP systems across a network that can translate documents, validate transactions, and preserve the commercial context behind them.

That work matters now because AI agents can generate actions faster than trading partners can reconcile inconsistent data. Procuros is pursuing the less visible infrastructure required for an automated order, shipment, or invoice to cross a company boundary without turning a formatting problem into an operational mistake.

About Procuros

Patrick Thelen and Benjamin Wulff met at university before encountering the same B2B integration problem from different operating seats. Procuros' company history says Thelen handled sales orders arriving by fax, PDF, and customer portals at a mechanical-engineering company. Wulff built interfaces connecting airlines at Airbus, where projects could take 1 to 2 years.

They founded Procuros around a straightforward premise: adding a new trading partner should not require another bespoke integration project. The company now reports more than 50 team members, more than €3B in transaction GMV, and €34M in total funding. Its backers include XAnge, W23 Global, Point Nine, Creandum, b2venture, and SquareOne Venture Capital.

Procuros serves suppliers that need to automate order-to-cash and retailers that need to connect long-tail supplier bases for procure-to-pay. Named customers include KoRo, Niche Beauty, The Quality Group, Tony's Chocolonely, followfood, flaconi, Chefs Culinar, Breuninger, and Hitschler.

The Problem Procuros Is Solving

B2B trade looks digital until two companies need their systems to agree. One organization speaks EDIFACT, another sends an Excel sheet, a third uploads a PDF, and a fourth requires a portal. Product identifiers, buyer codes, tax rules, payment terms, delivery dates, and quantities can describe the same transaction differently.

Traditional EDI moves structured documents, but every partner relationship can still bring mapping, testing, onboarding, and maintenance work. Small suppliers may lack EDI entirely. Large retailers may impose market-specific formats and tight deadlines. The operational burden falls on sales, supply-chain, finance, and IT teams that must decide which record reflects the actual commercial agreement.

Procuros turns that fragmentation into a network problem. A customer connects once, while Procuros translates EDI/EDIFACT, X12, XML, CSV, Excel, JSON, PDF, API, AS2, SFTP, and portal data between trading partners and ERP systems.

How the Procuros Platform Works

Translation is only the first layer. Procuros says every transaction can be checked against master data, contracts, and order history before it reaches the customer's ERP. The platform can match article numbers, GTINs, buyer and supplier IDs, order references, tax, totals, delivery details, prices, quantities, payment terms, and addresses.

The company supports orders, order responses, shipping notices, invoices, product catalogs, and master data. Its technical documentation describes a RESTful JSON API with an OpenAPI specification, while the main platform also exposes API, MCP, and portal paths for customers building automations.

The strategic asset is the context joining those documents. An invoice does not exist on its own. It belongs to an order, a shipment, a contract, and an agreed set of commercial terms. Software can extract a number from a PDF quickly. Acting on that number requires knowing whether it belongs there.

What Customer Deployments Show

Procuros and XAnge report more than 300 customers across DACH and the UK and more than €3B in annualized order volume. The company's published case studies show how the network model behaves inside operating teams, although the outcome figures are company reported rather than independently audited.

KoRo reports using Procuros across more than 300 trading partners, processing more than 20,000 invoices annually and reducing manual processing time by 90%. Niche Beauty says it digitized more than 300 active suppliers and brought 100 suppliers live in one week. The Quality Group reports connections to more than 30 major retail partners and thousands of EDI transactions processed each month.

Followfood reports 29 connected trading and logistics partners, more than 15,000 yearly transactions, and 5 to 6 hours saved per person each week. Tony's Chocolonely says a central API connection helped it extend automated document exchange into additional European markets with less internal IT involvement.

These examples span consumer brands, retailers, food suppliers, and different ERP environments. The recurring job is not glamorous: keep the commercial record clean while volume, partner count, and geographic complexity increase.

Why Procuros Matters for Autonomous Trade

AI agents raise the value of reliable transaction context because automation scales both good decisions and bad data. An agent may decide that an order can ship, an invoice can be paid, or a discrepancy can be resolved. Before that action reaches another company, the system needs to know which identifiers match, which terms govern the transaction, and when the evidence is weak enough to require a person.

Procuros is positioning its network as that trust layer. The company does not need every supplier and retailer to replace its ERP or adopt the same document format. It needs the data crossing those systems to become consistent enough for automation to operate within known rules.

The distinction matters. Fully autonomous supply chains are an ambition, not a verified condition across Procuros customers today. The company's evidence supports a live connectivity platform, customer volume, document automation, and increasingly rich validation. The next stage is proving that accumulated context can reduce exceptions while keeping human judgment available for the cases software should not decide alone.

Leadership, Culture, and Selective Hiring

Procuros describes a company organized around four principles: move faster than expected, do fewer things but do them properly, own outcomes fully, and win as a team. The language fits the product problem. Connectivity work is a graveyard for teams that move quickly around the details instead of through them.

The company's careers page also makes an unusual choice explicit: Procuros says it no longer hires publicly. It reports that referrals and the team's network produce better hires for its current stage, while leaving a direct channel open for exceptional people who feel aligned with the mission. There is no verified public list of open roles.

That policy should not be mistaken for a generic hiring pitch. It is a market signal about how a 50+ person company entering more of Europe and the United States intends to protect speed and cohesion while adding specialized talent. The trade-off is equally clear: a network-led model can miss strong candidates outside the network, a limitation Procuros acknowledges.

The Market Signal Behind Procuros

Procuros sits where enterprise integration, supply-chain operations, and agentic software converge. The company has funding, customer references, volume, and an expanding network. Its harder task is to make each additional connection improve the system without allowing complexity to return through the side door.

The company will be measured less by the number of formats it can ingest than by the quality of the decisions its infrastructure can support. If orders, shipping notices, invoices, master data, and exceptions resolve to one trusted commercial context, AI agents gain a foundation for useful authority. If they do not, the automation simply reaches the mistake faster.

That is the company Procuros is trying to build: the layer beneath autonomous trade that makes the next action safe enough to become real.