Animated GIF showing rotating images for this case study

Shaping an Expert System for Worker Co-ops

Year:

2023

CLIENT:

Graduate Research

Role:

Lead UX Researcher & Designer

Categories:

Expert System Design, AI/ML Product Design, Rapid Interface Prototyping, User Testing

How might emergent technologies support worker co-op development? How might those tools be informed and shaped by a culture of cooperation?

Building from their graduate capstone research, Puicón-Pérez investigated ways emergent technologies might help expand support systems for worker co-op startups. Through 2 expert interviews and review of recent AI/ML literature, the need for a configurable expert system emerged—one capable of supporting cooperation with integrity.

In response, Puicón-Pérez designed COOP•R, a multimodal proof of concept AI-powered expert system to provide tailored advice via text and voice. Early rapid prototypes were built using Figma over 4 weeks. Play testing with 5 participants provided validation and feedback.

Insights from Journey Maps & Interviews

Experience mapping and heuristics analysis conducted by Puicón-Pérez showed that people new to worker ownership often navigate dense resources alone through a disjointed and time-consuming process.

Conversations with a worker-owner and a co-op developer revealed two key insights. First, worker co-ops are shaped by their people and developers outside a group have limited firsthand understanding of challenges as they arise. Second, adopting a cooperative mindset takes work, especially when group members are overcoming obstacles like shame and trauma.

Considering AI/ML & Algorithm Design

Scholars ring the alarm that equitable design is critical for emergent technologies like AI/ML. In developing tools for worker co-ops, a human-centered, cooperative design strategy is necessary. Without it, AI/ML risks misrepresenting needs if not properly and ethically implemented.

To deepen the proof of concept, Puicón-Pérez imagined The Facilitator's Equation algorithm which aims to replicate facilitation, guiding potential AI/ML to match human-facilitated worker co-op development and support.

Iterative Prototyping & Playtesting

Over 3 weeks, Puicón-Pérez iterated the COOP•R interface 5 times using Figma. Impromptu play tests after each round surfaced insights that informed adjustments. By not providing guidance, informative user feedback emerged and helped further shape the experience. This rapid process helped achieve the essence of the AI-powered, configurable expert system.

Now, Puicón-Pérez is using Origami to build a more complete version of COOP•R, with the goal of engaging product engineers. Additionally, user testing with 10 participants on Maze will enable continuous feedback loops to iteratively improve the experience.

Featured Work

Considering Cooperation: A Design Framework

UX Research, Design Frameworks, Design Artifacts

Cultivating a Community Lab for Co-op Development

UX Research, Visioning, Service Design Prototyping, Play Testing

A. Puicón-Pérez

UX Research & Design

Connect on

Medium