Note: For reasons of confidentiality, I have omitted key information and real data when demonstrating my work and design process. What is shown here is a simplified case study that includes my learning takeaways.
Internship
Polymerize Labs is a B2B SaaS product, an AI-powered material informatics platform for companies in the material sciences industry.
The platform enables researchers to manage their data and make predictions through machine learning, cutting back on cost and time for R&D and driving the material innovation process.
Dec 2021 - Apr 2022
(5 months)
As product design intern, I redesigned several parts of the platform’s core data management features to drive greater usability for all users.
During my internship, I had the opportunity to design and handoff multiple features, champion a research project on design systems and collaborate closely in a cross-functional team of designers, product manager, materials engineer and developers.
My work revolved around two core components of Polymerize Labs: Work Orders and the Inventory.
The Google definition of a work order is: a document containing information about a task or project that can be scheduled or assigned to someone.
Users can create and manage individual projects that have multiple work orders, enabling better documentation and team collaboration.
A researcher might use a work order to consolidate their R&D and drive the development of a specific material or product. These work orders improve data analysis and allow researchers to inform their teammates on the process, results, and statistics of experiments.
The inventory is a database that stores parameters for existing projects and work orders. These parameters contain information such as category, properties, and unit of measurement.
Users can choose to enter data directly in the inventory. Data that has been uploaded or edited in other parts of the platform — such as work orders — is also stored and updated in the inventory.
The redesigns were driven by the pain points of existing customers:
I talked to my product team to understand our users, including a materials engineer who was closely familiar with the research process. I needed to better understand why researchers use Labs and what they need in the product, especially in work orders and the inventory.
Before, parameters stored in the inventory only had a single unit of measurement (UOM). Users could only use a single UOM throughout their work orders.
The properties of an ingredient can be measured in different units, for example, grams (g) or weight percent (wt%). If users wanted a property in another unit of measurement (e.g. litre), they would have to duplicate the property in the inventory entirely to replace the existing unit of measurement with the intended one.
This often led to many duplicate parameters. Users would then have to overwrite the units that were already used throughout all the work orders in the platform.
In their work orders, users were unable to duplicate trial data from one work order and create a new order consisting of data from those trials.
This was important, because there are cases where researchers rerun trials using the same parameters (ingredient, processing, or characterisation) and values, or rerun them and adjust a few parameter values to observe the changes in properties.
A key challenge for the duration of my internship in a B2B space was my limited access to primary research or user testing. I also did not have domain knowledge when it came to material sciences.
I worked around this by relying on good cross-functional team collaboration, past research or secondary research, and researching industry best practices in much of my design process.
My designs went through a few iterations based on feedback I gathered through multiple reviews. This consisted of internal design critiques in my design team, cross-functional meetings with a PM, materials engineer and feasibility checks with engineers.
I kept Nielson’s 10 Heuristics in mind as standards for evaluation for my designs, from sketching wireframes to creating the initial designs in Figma. By and large, they worked well as usability guidelines for Polymerize Labs, aligning with both the goals of researchers and the company vision.
After input from a design review, I switched to displaying the units of measurement as choice chips, instead of expanding the modal on click for a new Edit Units section. This gave users greater selection control, and ensured consistency as options in the Edit Data modal should be kept editable.
After consulting a few engineers, I removed the ability to add, select, and delete units in Work Orders, such that users can only toggle between units in their work orders. In addition, users can only delete units from the inventory that are not in use for any work orders.
Trials selected from the original work order could be displayed using pagination or infinite scroll. Infinite scroll was chosen since the content was located in a new modal.
I designed a batch renaming feature for the selected trials. This was shelved after conducting a feasibility check with engineers (but was noted as a promising feature for future product iterations!)
Adding, selecting, and deleting multiple units within the inventory (click to expand image)
Selecting trials from an existing work order (click to expand image)
Copying selected trials to create a new work order (click to expand image)
Switching between new and original work order
All my designs have shipped since the end of my internship.
In addition to delivering a solution that drives usage of the product by solving the problems of users, I realized that my designs also impact how users perceive the product and brand.
My designs also opened up opportunities for product improvements in future iterations, such as the batch renaming feature for trials and in other exploratory north-star work! (See below)
I also did collaborative/pair designing that explored opportunities for future product iterations.
Researchers have unique procedures and workflows. With multiple ongoing projects, it can be unproductive to keep setting up fields, parameters, and inputs to fit their procedures.
We designed a library that stores these “templates” that help them set up these workflows and reuse previous formats, that allows for template creation and sharing.
Here are some of our concept designs:
I also co-led an exploratory project on building and maintaining design systems for the company.
In addition to research on B2B design systems and industry best practices, we conducted a UI audit to gather buy-in from internal stakeholders, and to define key solution areas for future work on an internal design system for the product.
A huge takeaway for me was learning how to collaborate better with engineers. Involve engineers early on in the design process where possible; check feasibility early and often!
I worked hard to make my designs clear for handoff: annotating my designs a lot, accounting for edge cases, making sure interactions, error and validation states (e.g. input validation and notifications) were accounted for. It made me really happy to receive a positive comment from one of the team’s engineers — a small thing, but an indicator that my efforts put me on the right track ☺️
In reality, uncertainty is normal and we hardly design things from scratch.
I learnt how to work around the constraints of limited resources and design debt. Scoping and working with ambiguity taught me a lot about how to design to solve problems within constraints.
It was important to understand how my design impacted other teams when explaining my rationale and receiving buy-in during meetings.
Because I went in with little domain knowledge, I relied on close team collaboration to help me understand the problem space I was designing for. I learnt to use critiques and internal reviews to question my designs: gathering feedback and bouncing ideas off one another led me to uncover edge cases I had not thought of.
Before this internship, I had more experience with the exploratory stages of UX: user research, problem definition, gathering requirements, strategy and ideation. But I wanted to strengthen my skills in UI design.
My internship with Polymerize allowed me to grow immensely in my capacity as an end-to-end product designer and honed my ability to work in a cross-functional team.
