By Olivia Wu | oliviawudesign.com

ExplainaBoard is a web-based evaluation system designed specifically for NLP developers. It aims to boost innovation and efficiency in the AI development process by providing strong analytics of their model performance.

cover-2.png

Duration: 4 months

Role: Design Lead

Tools: Figma, Photoshop, ****React JS, HTML/CSS/JS

Collaborators: Jenny Wang, Lisa Leung, Ann Tang, Yo-Lei Chen


Background

ExplainaBoard is a powerful web-based evaluation system designed specifically for NLP developers. Its interface enables developers to upload their models and gain valuable insights through analytics. This system provides a detailed breakdown of model performance, highlighting strengths and areas that require improvement. With ExplainaBoard, developers can easily assess what their models excel at and identify specific areas that need further refinement.


Challenge

Navigating the Highly Technical Nature of the AI/ML Field:

One of the primary challenges we encountered was the complex and technical nature of the AI/ML field. To effectively address this, we undertook extensive efforts to continuously build our domain knowledge. Through 16 weekly client meetings, numerous co-design sessions, and 21 user interviews, we actively engaged in research to gain insights from users and the client.

As a product designer with a background in statistics and machine learning, I was able to contribute by simplifying and clarifying technical jargon, making it more accessible and understandable to design for.