IN Brief:
- Hardis plans two AI agents for Q1 2027 addressing documentation and warehouse performance analysis.
- Native WMS and OMS integration and yard management functions are already available.
- The roadmap also proposes more frequent software updates and wider supply chain orchestration.
Hardis Supply Chain has outlined a product roadmap combining more frequent warehouse management system updates with future artificial intelligence tools and closer coordination between warehouse, order and transport processes. The programme includes two AI agents scheduled for release in the first quarter of 2027. Existing integration between the company’s warehouse and order management products should be distinguished from those planned agents and from later proposals for automated operational decisions.
At the centre of Hardis’s existing software is the warehouse management system, which coordinates receipts, storage, internal movements and dispatch. Its records identify available stock and assign work to warehouse staff or automated handling equipment according to business rules. Order management systems work at a different level, deciding how customer demand is allocated to available facilities and following the progress of individual orders. Connecting the two can help prevent promises being made against inventory that is not actually ready for collection or picking.
That warehouse execution data is already connected to Hardis’s order management software, with yard management and carrier appointments covering adjacent activities. A warehouse may complete an order but be unable to dispatch it because the assigned vehicle is late. Conversely, an early vehicle arrival can occupy a loading bay when goods have not yet been prepared. Shared scheduling information helps operators recognise these conflicts and coordinate work around realistic transport arrival and departure times.
Building on those interfaces, the company intends to coordinate more of the movement between warehousing, orders and transport through a wider orchestration approach. This includes a planned portal offering a broader operating view, alerts about disruptions and tools to support collaboration between logistics partners. Such a system depends on consistent data across order references, vehicle movements and transport milestones. If connected applications describe the same event differently, operators can receive contradictory information even when each individual system is performing as designed.
The first planned AI agent, Solution Expert Agent, will provide natural language access to Hardis WMS documentation. Staff would be able to ask questions about product behaviour or configuration without navigating manuals through conventional search processes. The second, Operation Insight Agent, is intended to help analyse warehouse performance and identify areas requiring attention. Both are scheduled for the first quarter of 2027 and must not be presented as generally available tools in October 2026.
The two planned AI agents will need different controls because retrieving documentation and analysing live operational records involve different information sources. An assistant interpreting instructions needs accurate, current documentation and a way of distinguishing authorised guidance from obsolete material. Analysing warehouse performance involves transaction records, operating times, exceptions and data definitions. A reported delay can have several causes, including unavailable stock, equipment faults or missed transport appointments. A useful analytical system must account for these dependencies rather than simply identifying an unusually high figure in a dashboard.
Later stages of the roadmap envisage AI models making predictions and suggesting responses to operational problems. Hardis has used unexpected carrier delays as an example of information that could prompt revised work priorities. A delay may justify changing the loading sequence, but the decision depends on which orders are complete, whether alternative vehicles are available and how other customer commitments would be affected. The roadmap does not establish measured improvements from such functions because their development and deployment remain future work.
For a later phase, Hardis has described an ambition for AI agents to initiate selected operational actions under defined conditions. This creates a stronger requirement for permissions, traceability and human oversight. Changes affecting stock allocation or warehouse machinery cannot be treated like ordinary text recommendations: an incorrect action may create a discrepancy or interrupt a live process. Hardis has identified governance and security as central to that stage, without committing to a release date for broad autonomous execution.
Alongside those proposed AI functions, Hardis is changing how frequently it delivers updates to warehouse management software. Delivering enhancements more frequently may make new functions available sooner, but warehouses often operate with customised integrations and equipment whose controls cannot be changed casually. Updates may require testing against scanners, conveyors, robotics, label generation and external business systems. Hardis has not published a universal update cadence or guaranteed that every customer will move to the new delivery model at the same time.
The first dated AI milestones remain in 2027, whereas the WMS and OMS connections described earlier are already available. Existing WMS and OMS integration provides a foundation for the planned broader coordination, while the 2027 agents will extend access to product information and performance analysis if delivered as scheduled. The eventual commercial effect depends on data quality, permissions, the reliability of connected systems and controlled introduction into live warehouse processes. No performance gain across customers should be inferred from the roadmap alone.
