Background

👩🏻‍💼 Role: Workshop facilitator, UX researcher, and Designer

👩🏻‍💼 Role: Workshop facilitator, UX researcher, and Designer

⏱️ Duration: 5 months

💡 Context: Project as part of my Human-Computer Interaction Degree, I joined the project in its second iteration to improve it based on previous user feedback

📁 Skills: UX design, UX research, Workshop facilitation, Figma

About the project

Small and medium-sized enterprises oftentimes have issues with being understaffed and can lack the human resources that allow employees to have consistent one-on-one contact for knowledge sharing and communication.

→ Therefore, the need for a passive communication channel, that could fill in these gaps was identified. This led to the creation of an artificially intelligent ChatBot, that can provide employees with timely replies to their critical questions. However, one main concern was identified after the first iteration:

Problem statement

The users are often unable to trust the system and confidently use the proposed answers.

Proposed Solution Framework

After conducting extensive academic literature research I identified the topic of explainability which has been shown to increase trust in AI systems.

• 💡 More on explainability

Process

Design thinking in UX involves empathizing with users to understand their needs, ideating creative solutions, and iterating through prototyping and testing to ultimately deliver user-centered designs that address real-world problems effectively. It emphasizes a human-centric approach to problem-solving, prioritizing user experiences and perspectives throughout the design process.

Research

In order to find solutions for the concerns above mentioned we conducted 3 design and research workshops with designers, developers, and users, to answer the question of how can the user trust the system through the lens of explainability.

The Workshops:

• Designer Workshop: This workshop was conducted with designers to first conduct a heuristic evaluation and then ideate on how design can be tailored to increase user trust in AI decisions, and to collaboratively identify the possible features to strengthen users’ trust in the answers generated by the chatbot.

Developer Workshop Miro Board

• User Workshop: This workshop was conducted with users based on a participatory design approach to understand the needs of the user, evaluate early designs with the think-aloud approach, and collaboratively ideate on possible solutions.

Designer Workshop Miro Board

• Developer Workshop: This workshop was conducted with developers to first conduct a heuristic evaluation and then Identify how the algorithm can explain itself to increase user trust in AI decisions and to ideate on feasible features from the developer's point of view.

User Workshop Miro Board