02
Research
Design Audit: Outdated page UI with unclear information architecture
I started by conducting a design audit on current product.
I substitute myself in the role of a first-time user. Once I clicked in the platform, I found that it fills with abbreviations that I don't understand, which I soon felt like I cannot find anything that I need in here.
User Interview Insights: Users are having problems looking for data
I conducted user interviews with data users and admins, which validated what I saw in the design audit. Based on the interviews, we gathered the following insights:
I never found the data I want without searching for 5 times, the categorization makes no sense to me.
— Data user who constantly needs to request data
We initially designer the categories based on maybe 10 data entries, but now there are few hundreds and it doesn't make sense anymore.
— Admin member who was involved in the initial design
Research Insights
Summarizing the objectives of the design and the research conducted, we established three design directions:
Keep the interface clean to lower the entry barrier for new users
Redesign the info arch for page layouts that encourage users to request datasets
Restructure categories and tags to make datasets easier to discover
03
Goal 1: Keep the interface clean to lower the entry barrier for new users
For the first design goal - simplicity, I began by brainstorming two completely different ideas — option A being a dashboard for users and admin to access with minimized explore data assets section, while option B being a full data exploration page with a minimal summary for users.
In the end, option B was picked up by the team. The team agreed that Option A might create a cognitive burden for first-time users due to its heavy information load. In contrast, Option B offered a simpler overall structure and placed stronger visual emphasis on data browsing.
v.1. High-fidelity Prototype
Moving from low-fidelity to high-fidelity, I used a new design system adapted from Bosch's base design systems and branding colours; while also added new features, mainly filters and recommendations to make the experience more clear.
Goal 2 & 3: Redesign page to increase findability and request rates
For the other two design goals, I focused on designing UI elements to improve the findability of data, which includes cards, filters, search bars and more.
04
Final designs
Explore data assets
Through feature like Last Ordered Assets, user can quickly find the assets that everyone else looked for and ordered for the same project.
Search for data assets
User can choose either searching in the search bar with autocomplete suggestions, or filter assets by tags (typically a project or the department they are working under) to find the results they need.
05
Takeaways
Next steps
Over the course of three months, I designed and shipped 30+ screens for the full redesign of the Digital Assets Platform, including key flows such as the Data Explorer, My Data Assets, Onboarding for new data users, and the Data Approving flow.
For the Data Exploration flow, I focused on making navigation clearer and presenting metadata in a more accessible way. I wasn’t able to include an in-screen onboarding tutorial yet due to the complexity of the flow, but I see it as an important next step to support first-time users. I also see room to continue improving the filtering experience by exploring options beyond simple tag-based filters.
My learnings
Learning the technical domain
I took the initiative to build a better understanding of the product’s technical background, which helped me communicate more effectively with the development and data teams.
Gathering and aligning feedback
In each iteration, I worked closely with PMs, developers, and business stakeholders to gather input and adjust my designs so they could be both useful and feasible.
Balancing user needs
I tried to consider both new and existing users — improving onboarding for newcomers while keeping familiar functions for experienced users to ensure a smoother experience for everyone.