AutoMate AI
About
-
Product designer
Designed and drove the initiative of a new AI assistant, AutoMate, within the Takeoffs feature on Houzz Pro, enabling professionals to generate faster, more accurate measurements through a simple and intuitive UI, leveraging machine learning models to accelerate client delivery.
Problem
As AI-assisted takeoff tools become increasingly available, there’s a growing need for Houzz Pro to differentiate in the residential construction market by making advanced technology accessible through a simple, intuitive experience.
Results
Designed and contributed as part of the initial AB test release for 300 users on ConX and Houzz Pro. Received a lot of positive qualitative feedback from users via email and Intercom.
-
Significant adoption rate of 350 measurements per month
-
Number of WAU increased by 38%
Understanding the feature
Shortly joining the Takeoffs feature team, I reviewed the control version to understand the current takeoff experience. I found that professionals needed to manually draw shapes for each area, creating friction and slowing down the workflow.
Through discussions with product managers and engineers, we were able to review competing tools and identify ways to accelerate takeoff delivery and enhance our competitive edge.
The updates
I created a prototype to align with our leadership team to incorporate this into the roadmap. The concept was approved quickly, as it directly supported our goals of increasing feature adoption and user engagement.
We proceeded to AB test the new assisted mode, tracking user adoption rates and the number of active users for the tool measured as key success metrics.
What I learned
Understanding the tool
Applying user feedback
Coming into the project, I had no prior experience with the Takeoffs feature or the typical workflow of a contractor.
Gaining this understanding was important as it helped me design an assistant feature that aligned with real professional needs and complemented existing behaviors.
The AB test proved successful and generated valuable qualitative insights.
Iterating based on this feedback was an important learning experience, helping me refine the design toward a more seamless and faster takeoff experience that better matched user expectations.