Bonnie Voie Portfolio
Explore some of the ways we’ve supported clients and projects through different phases of growth and AI product development.
Start-up Strategy with Ripple Health AI
The founders of Ripple Health AI had a new vision for healthcare professionals to master communication skills: leveraging AI simulations and coaching to bridge the gap between costly, and unscalable in-person coaching, and impersonal, check-the-box training modules. I’ve been working with them to bring this vision to life.
Early on, we identified the traditional enterprise SaaS model as a risk, with healthcare budgets tightening even for solutions that offer significant cost-savings. Instead, we developed a flexible simulation technology, with feedback modelled after the co-founder’s own communication philosophy. We designed the product to meet potential customers where they’re already learning, instead of another SaaS platform with high adoption costs.
Instead of limiting Ripple to a B2B business model, we launched B2C continuing education courses using a flexible LMS platform that embeds our simulation product, integrating practice directly into communication training. Individuals and organizations enroll in Ripple training courses directly, and enterprise clients work with us to develop custom training curriculum with simulations that live directly in their organization’s learning tools.
A Strategic Reset for Literrific
The team of literacy researchers and Luxembourg-based engineers at Literrific spent two years building a powerful formative student literacy assessment rooted in the Science of Reading, that measures skills at the phonemic level (articulation of sounds). The product unlocks unparalleled scientific rigor and depth in the literacy space, but the team’s primary focus had been on the literacy analytics. The student assessment experience wasn’t engaging in a way that brought out best effort, and the educator dashboard was rich with metrics, but not actionable insights.
I work with Literrific to develop user experiences that match the innovation, technology, and the team’s passion. We introduced a modern literacy assessment experience that adapts to the student’s skills from pre-K to grade 6, with an encouraging mascot to guide students through the flow. For educators, we use the phonemic analysis to create actionable insights for both individual students and groups who share strengths and skills-deficits. The deep-dive metrics are still there, but the focus is on empowering educators with the data they need to plan instruction and intervention.
AI Advisor on SproochePrett
I developed SproochePrett to help learners of the Luxembourgish language overcome fears of speaking in public and get targeted corrections on their pronunciation and grammar. It meets an immediate need of helping learners pass the Luxembourgish Sproochentest, a requirement for Luxembourgish citizenship, while empowering them with the confidence and conversational fluency they need to integrate Luxembourgish into everyday life.
Luxembourgish is considered a low resource language, with linguistic documentation still developing. Many learning products use GPTs to help with grammar and pronunciation, but the underlying language models use German to “fill in the gaps,” leading learners down the wrong path. So despite its name, launching SproochePrett with AI support from the start was not the right approach. The current simulation offers learners the chance to practice using pre-recorded text-to-speech and stores learner responses without corrections. Now, I am engaging Luxembourgish teachers and researchers to collaborate on official language models and develop feedback processing loops so that when I release AI analysis and coaching, my users and I will be able to trust the quality of corrections.
