Bella
Building a personalized AI-native learning journey powered by a robot companion
An AI-native learning platform reshaping the edtech landscape through physical robot companionship
About the project
Bella is a groundbreaking learning robotics platform, building the next generation of AI-native education. Rather than simply teaching children about artificial intelligence, Bella lets them learn directly with it — through a physical companion that observes, interacts, adapts, and grows naturally alongside every child.
At the heart of the platform is Bella, a social robot. Children teach it through interaction, explore machine learning through experimentation, and receive personalized guidance across science, math, languages, creativity, and problem-solving. Combining a physical companion, multimodal interaction, and on-device AI, Bella creates learning experiences that feel conversational. Built offline-first for classrooms with limited infrastructure, it syncs when connectivity returns, bringing equitable AI education anywhere in the world.
.avif)
.avif)
Challenge
Quality AI-driven education remains out of reach for many schools due to limited connectivity and resources. The client set out to close this gap with a personalized learning companion, independent of infrastructure. Our team helped create an AI-native platform combining physical AI, adaptive learning, and offline-first deployment to deliver a tailored experience to every child, regardless of location.
Our process
To bring the learning platform to life, we combined strategic discovery with rapid, iterative execution across every stage of design and development, keeping the needs of children and teachers at the center of every decision.
Branding
Before translating Bella into the app experience, we established a strong visual foundation, building a brand capable of carrying the product while making it distinctive and trustworthy.
{{dropdowns-1}}
.avif)
UX/UI design for app and admin panel
Designing for a product this complex required balancing engagement with clarity. We focused on making every interaction feel effortless without oversimplifying the experience.
{{dropdowns-2}}
Website design
To communicate Bella’s value to the right audiences, we designed a landing page that presented the robot, platform, and classroom use case in a clear visual narrative.
{{dropdowns-3}}
.avif)
Development
The final stage involved implementing the designs into a fully functional, offline-capable app and a responsive website, ensuring uninterrupted classroom learning and a scalable brand presence.
{{dropdowns-4}}
.avif)
Results
- End-to-end digital ecosystem was delivered within one year
- Cohesive brand identity positioned Bella as a standout, credible player in the edtech market
- Child-first app experience made complex robot interaction feel effortless and alive
- Role-based admin panel gave teachers and admins full operational control without a deep technical background required
- Conversion-focused landing page helped communicate Bella’s value clearly to potential users
- Offline-capable PWA enabled uninterrupted learning through direct device-to-robot communication
- Production-ready Webflow site set Bella up to scale its web presence alongside the platform’s growth
Review
{{review}}
Our approach
Strong results start with open dialogue. At every stage, we work closely with the client’s team, running workshops, gathering feedback, and refining ideas through multiple iterations. If your product needs the same level of care, contact us — and we’ll help turn complexity into an experience people can understand, trust, and use with confidence.
{{approach}}
{{slider-1}}
5.0
“As AI reshapes our society, education must not simply adapt — it must lead this transformation. Bella is our contribution to that future: a vision for how the next generation will learn, create, and thrive alongside AI. Halo Lab has been an exceptional partner in bringing that vision to life, translating ideas into thoughtful design and intuitive software with care and diligence.”
.webp)
.avif)
.avif)
.avif)
.avif)
.avif)
