Dayli AI
My role sat between product thinking, implementation quality, and measurement.
The challenge wasn’t adding more screens. It was giving the product structure.
The Lovable prototype proved the idea could exist. The next step was figuring out how it should behave as a real product: where people begin, how they move through a need, what happens when the ideal path breaks, and how the team learns from real usage.
The interface could look right and still give the wrong answer.
Find a Solution already existed. My work focused on helping the team make the guided flow clearer, more resilient, and more trustworthy in the working product.
Start with the person, not the product category.
Not every user can describe a need through a structured flow.
Chat with Dayli AI was added as a new conversational entry point. I supported its definition, implementation quality, and QA alongside the guided Find a Solution experience.
QA became part of the product-design process.
Instead of checking only visual polish, I tested whether the working product still made sense when content, screens, and user behavior stopped being ideal.
I worked between product design and implementation.
Amber, our AI Lead, created the Claude Code-based development system. I worked within that system as a product designer, supporting implementation through prompting, code changes, scripts, review, and continuous QA with Dhivya, our Dev Lead, and Chelsea, our Project Lead.
Clarify the product behavior or gap before changing the build.
A recommendation is only as good as the data and logic behind it.
Supporting solution-fetching scripts helped me understand why product QA sometimes had to move upstream. When a recommendation was wrong, the interface could be functioning perfectly while the retrieval layer was not.
Shipping was not the finish line.
GA4 gave us a way to understand what happened after launch: how people moved across the two websites, which product experiences they used, and whether connected integrations were actually working.
GA4 let us see the relationship between the nonprofit’s main site and the product experience, including where visitors crossed between them and which paths actually continued.
The interface is only one layer of the product.
Dayli AI pushed my role beyond making screens. Product quality also depended on the data being fetched, the logic behind the interaction, the implementation holding up in real conditions, and whether we could measure what happened after launch.