When the industry talks about AI in manufacturing, the focus is still on the factory floor. Robotics, automation and production efficiency dominate the conversation. But one of the biggest inefficiencies in the industry today is not in production, it is in procurement.
While production processes have been optimized for decades, many procurement functions still rely on manual research, fragmented systems and limited visibility into supplier capabilities. As supply chains become more volatile and complex, this gap is becoming a structural bottleneck. Manufacturers are running highly optimized production lines, but making sourcing decisions with incomplete information.
Procurement under pressure
Manufacturers today face a combination of rising cost volatility, geopolitical uncertainty and increasingly fragile supply chains. Decisions that once followed relatively stable patterns now require constant reassessment.
At the same time, procurement teams are expected to move faster. Yet in many organizations, supplier discovery still depends on manual searches, siloed databases and Excel-based decision-making. Companies often lack visibility into available production capacities across regions, making it difficult to identify viable alternatives when disruptions occur.
When a supplier suddenly becomes unavailable, teams often need to restart the sourcing process from scratch, manually identifying and qualifying new suppliers under time pressure. Sourcing decisions that should take hours often take weeks. This imbalance between growing complexity and limited resources is one of the key reasons why AI is moving into the back office.
From production data to decision support
The role of AI in manufacturing is expanding beyond machines. It is increasingly used to support decisions that were previously based on incomplete or fragmented information. Modern AI tools can process large volumes of industrial data, including supplier information, machine capacities and market developments, and translate them into structured insights.
GieniExplorer, the market research agent built by Gieni AG, supports procurement managers with precise market analyses, customizable reports and an interactive interface to interpret complex market data. Such insights allow procurement teams to identify suitable suppliers, compare capabilities across regions and evaluate risks based on real-time data. Instead of asking “Who do we already know?”, teams can ask “Who is actually available and capable right now?”
What is emerging is not just automation, but a shift towards data-driven decision support.
A fragmented data problem
One of the core challenges in procurement is not the lack of data, but the lack of structured access to it. Relevant information is often distributed across different systems, formats and sources. Supplier data, production capacities and market signals are rarely connected in a way that enables fast and reliable decision-making.
Even when the right supplier exists, they are often not visible at the right moment. This fragmentation makes it difficult to gain a comprehensive view of the supplier landscape, especially at a global level. Platforms that aggregate and structure industrial data are beginning to address this issue, connecting manufacturers, machines and production capacities across markets.
From insight to execution
Better insights are valuable. But knowing what to do is only half the problem. The other half is doing it, drafting supplier outreach, preparing RFQs, coordinating internal approvals, logging interactions in the CRM, generating the report for management. The operational layer of procurement is still where most of the time is spent, even when the analysis is fast.
This is the gap GieniABX by GieniAG was built to close. ABX is the execution layer that takes the context generated by GieniExplorer, supplier shortlists, market analysis, sourcing benchmarks, and runs the workflow end-to-end. It doesn’t replace the tools procurement teams already use. It works through them, coordinating with the CRM, Microsoft Teams, email and existing approval workflows. And ABX workflows can be scheduled, so a weekly supplier risk briefing or a monthly capacity overview runs automatically once it has been configured.
A procurement manager can query alternative suppliers in Explorer, hand the shortlist to ABX, and have personalized outreach drafted, follow-ups scheduled, CRM updated and a status report prepared for the next sourcing review. The team approves and signs off. The work gets done.
Beyond automation: the rise of white-collar productivity
This shift is often described as “white-collar automation”. While automation in manufacturing has traditionally focused on physical processes, the next wave targets knowledge-intensive tasks. In procurement, this includes supplier identification, market analysis, risk assessment, and the recurring operational follow-through that turns analyses into completed work. The same structural pattern, structured knowledge work done by hand, exists across sales, finance and operations, which is why ABX was built as a general execution layer rather than a procurement-specific tool.
AI can process large datasets and detect patterns that would be difficult to identify manually. Emerging supplier clusters, capacity shifts, potential risks based on historical performance and external signals. This does not replace human expertise. It changes the role of procurement teams, from gathering information to interpreting, deciding and acting on it.
The effectiveness of these systems depends heavily on data quality and transparency. Many industrial companies still operate with fragmented data infrastructures, and AI-driven recommendations and actions must remain explainable to build trust. AI can support and execute, but it cannot compensate for fundamentally poor or incomplete data.
A structural shift in manufacturing
The discussion around AI in manufacturing often focuses on machines, automation and smart factories. But a significant transformation is taking place in the offices behind the factory floor. Procurement, market analysis and strategic planning are becoming more data-driven, with AI enabling faster and more structured access to complex information, and increasingly, the operational work that follows.
The next major productivity gains in manufacturing may not come from faster machines, but from better decisions, executed faster and repeated reliably. For a long time, competitive advantage in manufacturing was driven primarily by production efficiency. Increasingly, it will depend on how effectively companies manage and use information, and how quickly they can turn that information into completed work. Procurement is becoming a data problem and an execution problem at the same time.
