rose schmiedeknecht

home

about

gallery

contact

\ \

senti

the problem

Retail employees are expected to balance operational tasks with constant customer interaction— often under high-pressureworking conditions.Research revealed that:

overview

Senti is an AI-driven self-service application designed to reduce employee burnout in large-scale retail environments by automating routine customer interactions.

Focus— AI integration, service design, accessibility

This creates a feedback loop:overworked employees → lower-quality interactions → decreased customer satisfaction

  • Type: UX/UI, Product Design
  • Context: Academic project
  • Role: Research, Concept, Final execution
  • Tools: Figma
  • Customer service expectationssignificantly contribute toemployee burnout
  • High task volume reduces both storequality and service quality
  • Employees struggle to maintain emotional responsiveness while completing physical work

the opportunity

Most existing research focuses on improving human-to-human service interactions. Fewer solutions explore how technology can redistribute emotional labor rather than intensify it.

​

This project asks:

What if routine customer interactions could be offloaded—withoutremoving human support entirely?

concept

Senti is a mobile, in-store assistant that combines:

Rather than replacing employees, Senti supports a hybrid service model, where automation handles repetitive questions and employees focus on higher-value tasks.

  • A navigational interface (store map + product finder)
  • An AI chatbot trained in store procedures
  • Multilingual support for accessibility
  • A human fallback system for complex interactions

key findings

  • 106-person employee survey (Target, CVS, Walmart, TJ Maxx)
  • 3 in-depth interviews with retail workers
  • Desk research on emotional labor and service systems

Employees aren’t just overworked— they’reemotionally overextended.Reducing interaction load can improve both employee wellbeing and customer experience.

research & insights

Methods

Insights

Emotional labor (constant friendliness, responsiveness)is a major stressor

88%

of employees say customer expectationscontribute to burnout

81%

report reduced stress if customer interactionswere partially automated

design approach

1\ Prioritizing Common Tasks

The most frequent customer questions:“Where is this product?”“How do I get to this section?”These became the foundation of the system:Find Product + Store Map

3\ Hybrid Service Model

Research showed users still prefer humans for complexor emotional situations.Solution:Built-in “Call for Help” feature

Smooth escalation from AI → human support

2\ AI as an Interface Layer

Instead of forcing users through menus, Senti allows:Natural language input

Automatic navigation to relevant screens

Context-aware responsesThis reduces friction and supports different user behaviors.

4\ Accessibility & InclusivityMultilingual support for non-English speakers

Highly legible typography (Lexend, Inter)

Large touch targets and simplified layout forquick comprehension

core features

product finder

Displays product location on a store map

Includes real-world aisle imagery

Option for guided navigation

Out-of-Stock Handling

Suggests alternatives

Provides product reviews for informed decisions

Store Map

Simplified navigation to key locations

Designed for both casual shoppers and task-based users(Instacart, etc.)

AI Chat

Handles complex or open-ended questions

Can trigger navigation flows automatically

  • Cool neutrals and blues → trust, technology
  • Mint → clarity and modernity
  • Coral accents → warmth and emotional signaling
  • A single-column structure
  • Large margins and touch targets
  • High readability from a distance (kiosk/robot use)

The interface balances futuristic clarity with human warmth:

In my original wireframe, the orientation was landscape. I had some difficulties with keeping elements in consistent, predictable positions, so as part of my refinement I switched to portrait.

The layout uses:

visual system

link tofigma demo

reflection

This project pushed me to think beyond interface design and into service systems and human behavior. I was particularly interested in how emotional labor shapes user experience— not just for customers, but for employees.

If I were to expand this project, I would further prototype real-world interactions between Senti and users in a physical retail environment, refining how the system behaves inhigh-traffic scenarios.

  • Customers receive faster, more consistent support
  • Employees are freed to focus on operational tasks
  • Service quality improves through reduced cognitiveand emotional strain

final outcome

Senti creates a system where: