Lavar
An interactive reading platform that brings books to life. Built with AI agents and voice AI technologies, the platform offers an improved form of reading and engagement with a book that gives the reader contextual insights, external references, and the ability to enhance their experience engaging with content.
We’re building LAVAR, a voice-first interface to books and training content that transforms static materials into adaptive, culturally responsive conversations. Designed for hands-free engagement, LAVAR enables learners to speak to a book—and have it speak back—asking questions, surfacing just-in-time explanations, and guiding mastery through habit-based scaffolding.
Powered by Vapi’s voice infrastructure, LAVAR delivers personalized learning through authentic, diaspora-voiced narrators and a real-time knowledge graph that evolves with each interaction.
We use Vapi to:
- Stream real-time transcription into an adaptive dialogue engine
- Trigger function-calls for insights, goal-setting, and retrieval-augmented prompts
- Maintain memory across sessions using Pinecone embeddings
Our initial pilots (Q3 2025) will launch in U.S. Title I schools and workforce programs, with a public release targeted for Q4. We're actively exploring LMS integrations and responding to interest from African EdTech partners where voice-first learning can bypass literacy and infrastructure gaps.
To us, voice is more than an interface—it’s access. LAVAR turns any book into a mentor, guide, or coach—making expert instruction audible, adaptive, and equitable for every learner.
Latest Updates
Lavar wins the 2025 vapi Build Challenge
In June 2025, Vapi kicked off the Vapi Build Challenge with a specific question in mind: "what happens when you give hundreds of developers a production-grade voice platform?"
