SparkMinds. AI-powered Parent-Guided Child Development Platform
SparkMinds is a personalized learning companion for parents of young children, designed to support language and developmental growth through simple, everyday activities. By combining expert-backed frameworks with AI-driven recommendations, the app helps parents turn daily moments into meaningful learning experiences — building skills consistently, one step at a time.
One question at a time, with a calm progress bar — no overwhelming form.
Adaptive multi-select that shapes the AI’s recommendations downstream.
The model proposes 5 developmental goals tuned to the child’s stage — parents pick what matters.
A warm completion moment — onboarding ends with delight, not a dead-end.
The personalised activity feed — what the model learned, turned into a hands-on 30-minute moment.
How I worked through it.
Understanding the System
Instead of jumping into screens, I mapped the system: Parent input → AI recommendations → Daily actions → Feedback. This helped me understand how logic and user experience connect.
Designing Key Flows
I focused on designing a few critical flows first: Onboarding → Goal Recommendation → Selection · Daily Activity → Guided Execution → Feedback · Progress Tracking → Reflection → Next Steps. The goal was to keep each flow simple, intentional and low-effort for parents.
Balancing Simplicity & Depth
1 activity per day instead of weekly plans. Guided steps instead of long instructions — depth on demand, not overwhelm by default.
Designing Around AI (Not Just With AI)
AI wasn’t just a feature — it influenced the entire experience. I designed goal recommendations based on user input, adaptive activities based on feedback, and flexible flows where parents can adjust or regenerate. The goal was to make AI feel helpful, not overwhelming or invisible.
Iterating Through Feedback
Throughout the process, I continuously shared early concepts, refined flows based on feedback, and simplified wherever friction appeared. This helped move from idea → clarity → usability.
“LOVE LOVE LOVE the UI/UX vibe. Every time I see it, I feel calm and supported. SparkMinds as an “expert companion” comes through for me clearly. I can’t wait to share it more broadly and get feedback.”
What I’d carry into the next one.
For AI products, the onboarding *is* the model’s training data — design it like a feature, not a form.
Showing the user what the system is learning in real time is the single highest-leverage trust mechanic.