Warehouse Companion

Building a data center requires a complex chain of materials, suppliers, warehouses, shipments, and operational handoffs. Oracle did not manufacture every physical component used in these builds. Parts such as cables, server cases, racks, power equipment, networking components, and supporting infrastructure moved through a warehouse fulfillment process before reaching the data center site.

When a data center needed parts, it submitted a requisition request. The warehouse received that request, converted it into a work order, and assigned the order to one or more pickers. Pickers then collected the required items from warehouse shelves, brought them to a packing station or loading dock, packaged them, labeled the shipment, and prepared it for transport.

Warehouse Companion focused on this specific part of the flow: Pick, Pack, and Ship.

The goal was to give pickers a tool designed for the environment where their work actually happened: on foot, in aisles, around shelves, forklifts, pallets, boxes, labels, scanners, and time-sensitive orders.

The concept of Warehouse Companion began with my in-depth exploration of Oracle’s data center and warehouse operations. I visited an Oracle data center in Virginia in November 2025, met with the operations team responsible for data center infrastructure build-out, conducted recorded interviews, reviewed existing tools, and analyzed the workflows supporting supply movement from warehouse storage to data center delivery.

I also facilitated a Design Thinking-style exercise with the team. The intent was to move beyond known frustrations and draw on their practical experience to identify better ways of supporting the work.

To make the findings useful beyond the immediate project team, I turned the research into a strategic storytelling deliverable. In addition to the interviews, I recorded screen-capture walkthroughs of the existing warehouse software while operations experts demonstrated the workflow, and documented the data center and warehouse environment that shaped the work. I edited these materials into a video research documentary combining interviews, Design Thinking outputs, screen recordings, and environment exploration. The video helped socialize the findings with the broader UX organization and cross-functional senior leadership, including SVP and executive-level stakeholders. It also became one of the key reasons UX was invited into the subsequent week-long workshop to help reimagine the data center and fulfillment business.

In January 2026, I participated in a week-long workshop with Oracle data center specialists, warehouse storage experts, warehouse managers, and third-party contractors involved in data center construction. This gave me a broader view of the end-to-end process and allowed me to validate the Warehouse Companion concept against realistic operational scenarios.

For Warehouse Companion itself, I was the sole UX designer and owner of the user experience direction. I defined the mobile workflow, designed the Figma mockups, built the interactive prototype, and validated the concept with stakeholders who understood the warehouse environment.

The existing warehouse tools did not support the full Pick, Pack, and Ship workflow. The primary system behaved closer to a static inventory list than a workflow application. It was outdated, slow, desktop-oriented, and poorly matched to mobile warehouse work. Pickers relied on laptops in an environment where they needed to move through aisles, handle items, operate around equipment, scan labels, load pallets, and prepare shipments. Some parts of the process depended on third-party tools, spreadsheets, and manual coordination outside the main system.

Several pain points stood out during research:

• Pickers had little support for route planning through the warehouse.
• The system inventory did not always match physical inventory.
• Multiple pickers could work against the same order or overlapping items.
• Items could be picked more than once and left behind at packing stations.
• Packing required repetitive manual entry.
• Order state was not reliably visible across the process.
• Workers had to compensate for gaps in the software through memory, workarounds, and informal communication.

One example made the cost of the old process easy to understand. If a shipment contained 50 items and each item took roughly 30 seconds to enter or confirm, the worker spent about 25 minutes on data entry alone. That was time spent maintaining the system instead of moving the shipment forward.

The problem was larger than interface usability. The warehouse teams were compensating for missing workflow infrastructure.

Warehouse teams responsible for fulfilling data center component orders needed a mobile, workflow-aware tool that could support order ownership, item picking, barcode validation, exception handling, packing, shipment documentation, and readiness for transport.

The tool had to reduce unnecessary movement, make work status visible, prevent duplicate picking, help workers handle inventory discrepancies, and support the realities of physical warehouse work.

Warehouse Companion was designed around a simple operational promise: help pickers know what to work on, where to go, what to collect, how to verify it, and how to prepare the finished shipment for the next stage of the supply chain.

The concept organized the picker’s work into a guided mobile flow:

Assigned order → Order review → Start picking → Guided navigation → Item validation → Quantity confirmation → Exception handling → Packing → Shipment photo → LPN generation → Completion

The design treated assignment, status, validation, and shipment tracking as part of the same experience. That was a deliberate shift from disconnected tools toward a continuous workflow.

Warehouse Companion was also a practical use case for AI in enterprise software without turning the experience into a chatbot.

AI supported the picker by optimizing the navigation path through the warehouse and recommending the most efficient order of items to pick. Instead of forcing the worker to plan the route mentally, the application could use the active work order, item locations, picker location, priority, and warehouse structure to guide the picker through the most efficient sequence.

The AI value was embedded directly into the work. It reduced unnecessary movement, supported faster picking, and helped the picker make better use of time without requiring a separate conversational interface.

Because I left Oracle before final implementation, these are not post-launch production metrics. They are based on research findings, timed task observations, prototype validation, and modeled workflow improvements.

90%+ faster item entry during packing
Timed research showed that manually entering each item into the legacy system took approximately 30 seconds per item. Warehouse Companion replaced most of that work with barcode-driven entry. Scanning took under 2 seconds, with manual code entry available as a fallback when scanning failed.

For a 50-item shipment, the legacy process could require roughly 25 minutes of data entry. With barcode-driven confirmation, the same item-entry task could be reduced to approximately 2 minutes, depending on scan conditions.

20–50% faster work order completion
The previous picking process had little meaningful route planning. Pickers had to search, backtrack, coordinate informally, or work without knowing whether another picker was duplicating part of the same effort. Warehouse Companion used AI-supported route optimization to recommend the order in which items should be picked and guide the picker through the warehouse more efficiently.

The estimated 20–50% improvement depended on order size, item distribution, warehouse layout, and how much unnecessary movement the old workflow created.

Inventory updates reduced from 10–20 minute delays to near real-time
In the old workflow, the system often was not updated until the picker returned to the dock and deposited the order. That created a 10–20 minute gap between physical inventory reality and system inventory. Warehouse Companion allowed item-level confirmation at the point of pick, reducing that delay to near real-time scan-based updates.

For this portfolio case study, I rebuilt the Warehouse Companion prototype as a portfolio-safe interactive Figma demo. The embedded prototype shows the full Pick, Pack, and Ship workflow from assigned work order through picking, exception handling, packing, shipment photo capture, LPN generation, and work order completion.

Warehouse Companion produced an implementation-ready mobile workflow concept and interactive prototype for Oracle’s warehouse modernization effort.

I left Oracle before a final implementation decision, so I cannot claim production launch or shipped metrics. At the time of my involvement, the concept had been reviewed with data center and warehouse stakeholders, tested against realistic operational scenarios, and positioned as a candidate for implementation.

The work demonstrated how a fragmented, desktop-oriented operational process could be reframed as a mobile workflow with clear ownership, visible status, item validation, exception handling, shipment proof, and AI-supported route optimization

Warehouse Companion reinforced a principle I consider essential in enterprise UX: complex work does not become simpler because software exposes more data. It becomes simpler when the product understands the workflow well enough to guide the user through the right action at the right time. The project required domain research, operational analysis, stakeholder facilitation, mobile UX design, and careful attention to edge cases. It also showed a practical use of AI in enterprise software: embedded workflow intelligence that helps people complete physical work more efficiently without forcing them into a separate conversational layer.

My effort produced an implementation-ready mobile workflow concept and interactive prototype for Oracle’s warehouse modernization effort.

The work demonstrated how a fragmented, desktop-oriented operational process could be reframed as a mobile workflow with clear ownership, visible status, item validation, exception handling, shipment proof, and AI-supported route optimization. For me, the value of the project was in turning a difficult operational process into a clear product direction: one that respected how warehouse work actually happens and gave pickers a focused tool for getting the right shipment ready, correctly and on time.