Almanden Genomics
Almaden Genomics is a science product and research team that began in IBM's Almaden Research Center in Silicon Valley. Comprised of many scientists and engineers, our team understood the struggles and complexity of creating bioinformatic pipelines. We believed that we could simplify this process by creating a user-friendly no-code workflow builder. With our application, researchers and scientists without much engineering knowledge of writing custom executables, can now to create complex bioinformatic pipelines quickly and more efficiently. In the end, my team and I were able to successfully take this application from pre-alpha to go-to-market in a matter of 11 months.
Current Process
Currently, bioinformaticians and engineers create their own custom tools and pipelines by writing code in languages such as WDL, Python, Java, and Perl. Overtime, these pipelines grow with more steps and parameters, making it more difficult to maintain these scripts. This puts an unnecessary amount of technical burden on the bioinformaticians to not only learn and understand how to create pipelines, but to also continuously maintain and update them.
Design Process
The first step was to understand the users. My team and I conducted stakeholder interviews in order to have a better understanding of their daily tasks. From these user interviews, we created user personas, user flows, and user journey maps for all stakeholders. This helped us empathize with our users and gave us a better understanding of user pain points. From this information, we were able to brainstorm, organized gathered data, and create preliminary sketches of the information architecture.
PAIN POINTS:
Technical Limitations
Resource Management
Error Prone
Design Studio
With these user personas in mind, we started visualizing the layout of the application. We began to break down the information architecture and create wireframes for those sections to showcase our ideas. Each team member sketched several ideas for each section of the application, then we voted on the features we liked and incorporated those ideas in the project wireframes. The goal of the application is touch upon three major pain points: Technical Limitations, Resource Management, and Error Handling
Wireframes
Technical Limitations
Scientists and researchers face a lot of unnecessary technical burden throughout their daily tasks. They are forced to learn how to code and maintain pipelines in order to run tests. Throughout this process, users will have to create and upload their own custom tools to manipulate data as they see fit. Creating and maintaining pipelines and tools takes months to learn and perfect. Our application relieves them of this burden by providing them with a drag and drop canvas, the ability to create and/or upload custom tools, and provide users with help.
Resource Management
Users also struggle with collaborating with other users, navigating large amounts of information, and tracking and managing changes to software code. Our application gives them the capability to share data, customize their workspace, and maintain code and tools w version control.
Error Prone
Since scientists and researchers need to learn how to create and maintain pipelines, users frequently run into issues while executing their pipelines. When these errors occur, users must identify the issue and solve it, which can take an extreme amount of time and effort. Our applications provides users with error notifications and preliminary warnings, and also allows users to lock specific items.
Challenges
- Losing a team member, forcing the team to pivot and continue to deliver a quality product within business time constraints
- Understanding and digesting complex subject matter while engaging in conversation with subject matter experts
- Move product from pre-alpha to go-to-market
A major challenge that I faced was having a solid understanding of the bioinformaticians complex tasks. One such example was creating parameters within a custom tool. I held many sessions with scientists and engineers in order to gather information to understand this extremely complex process. Once I had enough information, I began constructing a flow chart showcasing the options users have in order to create a parameter.
After a few brainstorming sessions with the design team, we decided to go with a wizard approach. This wizard approach allows users to focus better on the content pertinent to each step and decrease the chance of errors. In our wizard approach, we give users the ability to either enter command line code or use a user-friendly interface to create parameters. If the user utilizes the GUI, they are prompted with five required fields. Once the required fields are filled in, more fields are revealed based off of the users pervious answers. This method makes it easier for the user to create a parameter without any errors. If the user inputs command line code, the parameters will be created automatically. Challenges like parameter creation made this project difficult, but these challenges yielded results that made our application more efficient and user-friendly.
Project Details
Role: Senior UX Designer
Timeline: 12 months
Takeaways
- Visual aids, along with documentation, will make handoff to engineering more efficient
- Frequent design critiques help streamline asynchronous design tasks
- Analog sketching removes designers from technical constraints, thus speeding up the design process