IN Brief:
- IFS Softeon's Fulfillment Flow uses specialised AI agents to analyse warehouse operations and recommend responses to changing conditions.
- Digital workers cover inbound activity, wave management, labour, automation, slotting and replenishment, and outbound operations, sharing information with a central agent.
- Warehouse managers retain final decision-making authority, while independently verified productivity improvements and named customer deployments have not been disclosed.
IFS Softeon has introduced Fulfillment Flow, an AI capability designed to coordinate decisions across connected warehouse activities, using specialised software agents to assess operational conditions and recommend actions to warehouse managers.
Announced at IFS Unleashed in Orlando on 13 October 2026, the system combines information from warehouse processes with a central operational agent that considers how decisions in one area may affect activities elsewhere in the facility. The approach is intended to address the limitations of optimisation tools that concentrate on individual operations without accounting for their relationships with the rest of the warehouse.
Fulfillment Flow organises its AI agents, described by the company as digital workers, around six operational areas: inbound, wave management, labour, automation, slotting and replenishment, and outbound. Each agent examines the conditions relevant to its area, while sharing observations with a central agent responsible for assessing the overall flow of work.
Warehouse processes are closely connected, even when they are controlled through separate systems or measured against different performance indicators. Goods received from suppliers must be checked, recorded and placed into suitable storage locations before they can be allocated to orders. Picking, packing and dispatch then depend on inventory availability, labour and handling equipment operating within the required time windows.
A delay in one activity can therefore alter the requirements of another. Releasing additional picking work may increase immediate productivity, but it can also create congestion at packing stations or sortation equipment if those operations are already approaching capacity. Conversely, holding back picking work may leave labour underused when downstream capacity becomes available.
Wave management illustrates this relationship. Orders can be grouped according to dispatch deadlines, inventory availability and the resources needed to complete them. The appropriate size and timing of a wave depend on the capacity of the picking operation as well as the activities that follow it. Optimising only the release of orders can result in work accumulating elsewhere in the facility.
IFS Softeon’s system is designed to consider these interactions when making recommendations. Its digital workers use a common framework covering the physical facility, live operating conditions, adjacent processes, site goals, available employees and automation resources. Information from the different areas is brought together by the central operational agent rather than being treated as a series of independent decisions.
The company specifically includes relationships with receiving yards and manufacturing line feeds within the system’s operational context. Both can influence warehouse priorities because inbound vehicle arrivals and production requirements impose constraints that are not always visible within individual storage or picking processes.
Slotting and replenishment present further examples of linked decisions. A warehouse may place frequently selected products in accessible picking locations to reduce travel, but those locations require replenishment as inventory is consumed. If the replenishment process falls behind, picking performance can deteriorate even when sufficient stock remains available elsewhere in the building.
Labour allocation introduces another variable. Employees may have different equipment qualifications or be assigned to tasks that must be completed before a departure deadline. Moving workers to relieve one backlog can reduce capacity in another area, making the consequences of a reassignment dependent on the current balance of work across the operation.
Fulfillment Flow is intended to bring those factors into a common assessment, using operational data to identify constraints and suggest possible responses. Warehouse managers can ask questions in natural language and receive explanations of the conditions contributing to an issue, together with recommendations based on the information available to the system.
The accuracy and usefulness of those recommendations depend on the quality of the underlying data. Inventory records, labour availability, equipment status and order priorities can change repeatedly during a shift. If information is incomplete or no longer current, a recommendation based on it may fail to reflect the actual operating conditions.
The announcement does not specify the system’s data refresh frequencies, required interfaces or procedures for validating individual recommendations. These details will be relevant when assessing integration with existing warehouse management systems, automation controls and operational reporting arrangements.
IFS Softeon has adopted a human-in-the-loop model, in which specialised agents submit observations and proposals to the central agent. The warehouse manager remains responsible for determining priorities and approving the response, rather than handing final authority over operational decisions to the software.
Mark Fralick, chief technology officer at IFS Softeon, said the approach is intended to bring warehouse context into decisions made during everyday operations. The company says Fulfillment Flow will follow management decisions and continue learning across shifts, although it has not published technical details of that learning process or independent evidence of the resulting performance.
Retaining managerial control allows decisions to incorporate information that may not be represented in the system, including unusual customer requirements, local operating restrictions or maintenance arrangements. It also establishes an approval step between identifying a proposed improvement and changing the allocation of warehouse resources.
The capability is being introduced within IFS Softeon’s existing warehouse management portfolio, following the acquisition of Softeon by enterprise software group IFS. The announcement does not identify a named customer deployment, a general availability timetable or published pricing, and no measured improvement in throughput, labour productivity or dispatch reliability has been disclosed.
The launch consequently establishes the system’s intended architecture and method of supporting operational decisions rather than proving its performance in a specific warehouse. Implementation evidence will be needed to assess how reliably its agents identify constraints, how managers use their recommendations and whether the resulting changes improve performance across the complete fulfilment process.
