The team spent 3 months researching the problem space thoroughly and defining key problem areas to tackle (education, collaboration, workflow efficiency), and another 3 months to test out solutions and building the interactive prototype and other service and design recommendations.

TL;DR

Research Methods

  • 20 Literature Reviews
  • 19 Expert Interviews
  • 4 Contextual Inquiries
  • 10 Semi-Structured Interviews
  • 20 Usability Testings
  • 2 Wizard of Oz
  • 10 Speed Dating Sessions
  • 5 Competitive Think-Alouds

Research Participants

  • Aspinity Internal Teams
  • Aspinity Client Contacts
  • LinkedIn Recruitment
  • Personal Network
  • CMU Alumni Network
  • Academic Researchers

our process

Hover to see a brief summary of each phase of the process!

key Research Artifacts

Interrelationship of engineers in the AnalogML process

AnalogML Application Development Workflow

How Astrl provides an ecosystem of support

User journey of an AnalogML Engineer

Learning Highlights

AnalogML is a new niche within the domain of ML. Currently, clients sit down with experts from Aspinity to learn the intricacies of the technology to develop a fundamental grasp of AnalogML. To empower clients to create their own AnalogML applications, client engineers need to understand the high-level relationships and requirements of working with analog data and hardware.
The end-to-end application development process involves multiple roles and expertise across the team. No individual knows how to do it all - yet everyone needs to be aware of each step of the process because different steps require specific skill sets with handoffs to one another. Effective team communication collaboration between diverse individuals needs to be aided.
Currently, there are many frictions and inefficiencies in how Aspinity engineers develop an application because tasks require lots of manual labor. Several steps requires the engineer to restart processes every time they make a change. Sometimes lack of context requires them to do research and experimentation. There is an opportunity for automation or smart recommendations that would ultimately save time.
Niche Domain Knowledge with Steep Learning Curve
Cross-Functional Teamwork Lacks Effective Collaboration
Manual Tasks are Inefficient, Inflexible, and Inevitable
View Solution