Overview
All products are designed based on customizing the main product XDP. A total of 3 integrated apps are showcased in this case study. Some prototype designs are illustrated at the end of this page. I worked with the team for 6+ months over 2 years.
Key Design Note: I not only worked on integrating those apps into the main product but also customized the products according to specific client needs and business requirements (only incorporating the critical features).
Product Documentations
Clients from various industries have different data analysis needs.
All customization projects start with small adjustments to the main product's project workflows due to the variability of clients' industries.
Case Study: 3 Integrated apps for bioinformatics and healthcare clients.
DaaS: data analyst-centric query and cleaning tool.
Bioinformatics tool: RNA sequencing analysis tool.
Goldfinger: an AI annotation tool for structured medical records.
Generally, the platforms allow users to create data analysis projects by:
1. importing their own private datasets
2. Importing from datasets that are authorized for use by others' private datasets
3. importing public datasets from the data marketplace
Then users use whichever integrated app they need to use for their purposes.
(App #1) DaaS Product Design
All of the integrated apps are located on a dashboard of the application marketplace, and users can add whichever app they need to use to their projects on the app marketplace.
I drafted logic and workflow to design features given customer inquiries. This is an example of data set authorization and corresponding payment logic for the platform. Since this product is for a specific buyer, the features must be designed strictly according to user needs. For this feature, my responsibility is to make sure payments are processed correctly according to project type, project scale, and resource usage.
Established corresponding workflow for the DaaS platform. Here is an example of the general catalog and pages. Those flowcharts are discussed and aligned with international client product managers and the sales team on a daily/weekly basis.
DaaS User Flow (Step 1)
DaaS enables users to analyze using some preset features in assistance mode or program themselves by using developer mode.
Left side: Imported dataset details
Users in developer mode can convert data requirements into query statements and run the datasets.
DaaS User Flow (Step 2)
Query results are shown below after query processing. Users can then add the processed query results to the project they are working on.
DaaS User Flow (Step 3)
Users can further execute their data tasks within the docker runner, a contained environment.
Designed Features:
Dataset catalog
Job terminal live display
System catalog
Managing output files
(App #2) Bioinformatics Tool
Users need to configure specific node information and project parameters, and then start the RNA-seq alignment tool to perform data analysis.
Bioinformatics Pipeline Integration
Users need to configure specific node information and project parameters, and then start the RNA-seq alignment tool to perform data analysis.
Start RNA-Seq workflow
In RNA-seq alignment, users select the input datasets to start the pipeline. There are limitations of the app itself, so a user-friendly alert (warning message) is provided on the top regarding file types.
Configuration
Users then configure specific workflow parameters, application settings, purchasing options, and AWS instance type, etc. All node information will be available in logs. Finished tasks would be available for export.
Composing the Use Case Tests
Here is a typical test case for RNA-seq data analysis. I worked with a bioinformatics engineer and the client's product manager to compose the use case testing documentation. Those use cases are first used for internal testing and then used again while customers receive the latest version updates.
Goldfinger is an application that automatically annotates medical reports and outputs standardized files. Users need to create a new project first.
Goldfinger is an application that automatically annotates medical reports and outputs standardized files. I prototyped the UI, so that the interface design is the same as the main product XDP, so the user experience can stay consistent.
Users can always manually double-check to ensure the highest annotation accuracy.
Medical Imaging
If Imaging is involved in a medical report, then users can choose to include specific data by annotating the images or adding comments.
After finishing and submitting the results to the cloud. Then users can export their entire project, consisting of numerous standardized medical records. The product enables users to further data analysis in-platform, for instance, using DaaS or RNA-seq alignment above.
Prototyping
Data Marketplace Initial Prototype
How I prototype those highly specialized/technical products:
Customer Focused: Since XDP products are for specific buyers (i.e., bioinformatics researchers), I would first interview bioinformatics engineering to understand their work, identify the main user needs (In this case, bioinformatics researchers working with RNA data), the most useful tool (RNA-seq alignment) and their data analysis needs.
Usability: XDP products are highly technical, and users will have a learning curve. I always ensure the technical 3rd-party built-in tools are localized to the same design language as XDP. For example in Goldfinger, users can configure and monitor their jobs in the exact layout and workflow as any data project on XDP.