Transforming machine learning concepts into solutions
Create, develop a roadmap for design
Interaction Planning (Wireframes/prototypes..)
Measure and analyse the effectiveness of the design
Collaborate with Machine learning PM, engineers, and data
ABOUT PROJECT
As the sole product designer, I redesigned the platform for AI model training, addressing significant usability concerns and rebuilding the visual appearance. By envisioning the experience of our data scientists and customer support team perspective, I enhanced the design, resulting in a 68% decrease in task completion time and a 139% increase in users’ subjective satisfaction.
THE CHALLENGE AND OBJECTIVE
Our engineers were under significant time pressure to deliver the MVP, so they had to compromise the user experience (UX). Subsequently, I conducted some rapid usability tests and identified critical usability concerns that led me to create two personas.
USERS AND PAINT POINTS
Automates all steps in document processing and complex tasks (classification, splitting, capturing, validation & enrichment) with No-code software and instead of being swamped with repetitive invoice completion, Hypatos studio users can spend their time more productively.
Advanced Users
Usage Inefficiency: Advanced users face inefficiencies in using the system as they constantly need to navigate across various pages to search for information amidst mixed data.
System Status Visibility: Advanced users tend to make mistakes due to the low visibility of the system’s status, which makes it easy to lose track of individual training progress.
Basic Users
Incomplete Understanding of Competency: Basic users need a more explicit mental model of competency and the essential information they need to collect.
Unclear Training Process: Basic users face difficulties in the training process as they are unsure what actions to take next.
QUESTIONS?
“What strategies can be implemented to clearly convey the training process to users?”
“How can we enhance user proficiency in the training process?”
“What approaches can we take to improve user understanding of the system’s status?”
PROCESS
I used a rapid cycle of paper and digital prototyping, user testing, and feedback from AI engineers. With each iteration, I gained valuable insights that enabled me to make various improvements to the design. Some of these insights helped me make minor usability tweaks, while others led to significant changes in the design direction.
DESIGN APPROACH
As users are my first concern, I decided to design a coherent training process in early iterations. Following several brainstorming sessions, I explored various methods of guiding users through the training process, such as adding steps, guide sections, guide pages, lots of tips, and navigation bars.
DESIGN DECISIONS
I planned a practical idea to create an overview page that provides users with a concise summary of all the information on the platform.
By presenting information this way, users can quickly locate the content they are interested in without navigating through multiple pages.
This approach not only simplifies the user experience but also improves the platform’s information architecture, making it easier for users to find what they need quickly and efficiently.
In essence, the overview page serves as a one-stop-shop for users to get a comprehensive understanding of the platform’s content and keep everything on track.
Overall, implementing an overview page is a perfect solution to enhance the user experience and streamline the training platform navigation.
After careful consideration, I decided to develop a training guide page with a dedicated work area. This approach allows users to navigate the iterative training process without the need to switch between multiple pages, thus minimizing any potential disruptions to their learning experience.
VISUALS
Next Case study
Web APP
Hyboard - Design System
Customisable Theme-Based Design system crafted to Elevate the Consistency and Quality of Digital Products through Hypatos platforms