
Due to an NDA with Epic Systems, I can't share my actual designs — the visuals shown are recreations I built independently.
I designed the OCR-driven packing slip verification feature on Epic's Rover app, letting procurement teams reconcile thousands of hospital shipments against POs and catch discrepancies in real time — boosting both speed and accuracy for loading dock workers.
I led user research and owned end-to-end design for the feature's workflow, then shaped receiving-workflow standards across Epic's Supply Chain suite — with this project adopted as the reference model going forward.
Accurate, efficient shipment logging is critical to hospital inventory management and vendor discrepancy resolution. Central to this process is the packing slip — a physical document itemizing package contents, which workers must manually compare against the corresponding Purchase Order (PO) to confirm the correct items were received.
Today, that means manually reading the packing slip, entering its contents into a computer, and tracking down the matching PO — a tedious, error-prone process that causes receiving delays and inventory discrepancies.

To understand the current receiving process, I mapped a sample workflow using notes from previous loading dock visits.

I then analyzed dozens of packing slips, annotating what counted as "important" information versus what could be ignored.
Since formats varied widely across vendors — no two followed the same structure — I focused on identifying common patterns and categorizing data by importance.

I also conducted two field studies on Epic's campus — at the Facilities Dock and Culinary Receiving Dock — each revealing a very different process.

Lastly, I led a scenario-based interview with the Director of Receiving at SUNY Upstate Hospital to understand the receiving flow specifically within a hospital setting.
“Unit of measure is a MONSTER between vendors”
“Person-to-person inconsistency is a common occurrence”
“Packing slips can contain information not detailed in the PO”
My developer partner built an MVP prototype to experiment with the backend OCR processes.

This helped me understand how the backend functioned, but the workflow still needed to be redesigned around what users actually needed.
Using insights from research, I sketched and iterated through each screen and interaction, grounding every design decision in a specific research finding I could point back to.



Open the scan feature, align the image, and capture the packing slip.
This creates a new shipment and displays an itemized list of the shipment contents.
Access the saved image anytime, eliminating the need to physically store the paper copy.
OCR errors are indicated with a warning symbol and stored as null values.
Item-level issues are flagged with a comment, signaling a need for further procurement action.
Epic's Supply Chain features are still being standardized ahead of a 2029 release. My research is informing that standardization effort, and my designs are serving as the bridge between procurement and payment flows.
The packing slip tool is projected to cut loading dock receiving time by 25%, with faster discrepancy catching.
Digital packing slip storage removes the need to file and retrieve hundreds of paper slips by hand.

Between shifting requirements, conflicting feedback, and endless iterations, it was easy to get lost trying to accommodate every edge case.
Returning to a core design principle — "good defaults by design" — helped me stop chasing every "if" and focus on what actually mattered.
Shipping in 11 weeks meant advocating for design decisions to shape the feature, not just fit around backend functionality.
Leading every field study and interview myself gave me the evidence to explain and defend each design choice with full ownership.