Pepsi Image Recognition Insights Tool

On this project I highlight my involvement in the development of Pepsi's Image Recognition Insights (PEPIRIS), an internal tool aimed at enhancing the efficiency of Pepsi representatives in executing the "Perfect Store" concept, ensuring optimal product display compliance.

The Task

My role encompassed a comprehensive overhaul of the PEPIRIS application, leveraging cutting-edge technologies such as image recognition, artificial intelligence (AI), and machine learning (ML) to significantly reduce the time spent by representatives at each store by nearly 50%.

Conducting a thorough heuristic analysis of the app, I scrutinized both original designs and production versions, synthesizing user feedback to identify critical experience pain points. Among the key recommendations was the simplification of template pages to enhance user understanding and streamline navigation. By remapping primary flows and restructuring interface elements, we laid a robust foundation for future feature expansion.

Specific attention was devoted to refining the image recognition functionality, a focal point of user frustration. Through iterative design iterations, we addressed usability concerns stemming from unclear button layouts and ambiguous visual cues. By enhancing the clarity and intuitiveness of the user interface, we facilitated smoother interaction and empowered representatives to leverage image recognition capabilities with confidence.

Racks

Each photo thumbnail (A) depicting the display rack now functions as a tab, facilitating seamless navigation and focused work on individual displays (B). In addition to showcasing the rack type, we incorporated the planogram type to preempt errors and accommodate adjustments as needed.

Clearer Image

We discovered that despite the quality of product photos, they weren't optimal for user interaction. Collaborating with the development team, we implemented an image swap for each product, resulting in a significantly clearer interactive interface while maintaining a swift reference view of the rack and planogram.

Clearer Interaction

Previously, when users needed to take action with a product, we utilized colored bounding boxes (red, orange, gray, or green). However, this approach garnered substantial negative feedback, as users found it challenging to discern the meaning of each color.

Our redesign enabled us to simplify and clarify micro-interactions, enhancing the overall usability of the interface.

Here are the completed Store Rack screens, marking a significant advancement for the PEPIRIS tool. Thus far, feedback from both our user base and partners has been overwhelmingly positive, indicating a step forward in usability and functionality.

The interface for capturing and analyzing images underwent visual refinement, featuring a significant enhancement: the incorporation of a progress bar. Despite its modest size, this addition yielded substantial benefits. Users lauded the inclusion of the progress bar, as it provided them with a clearer indication of when the process would conclude, enabling better task planning and smoother transitions to other activities.