All about Vocina

Scarlatti have developed Vocina – designed and built in New Zealand – to support learning and assessment across a range of educational settings. Our oral AI agents help educators save time, support students through scaffolded learning conversations, and strengthen the evidence for grading and moderation. We also offer assessment approaches that students often find more engaging than traditional formats.

Our catalogue of oral agents include:

• Viva Vocina. A short, viva voce-style conversation that helps learners explain and discuss submitted work, demonstrating understanding and reasoning.
• Formative and summative oral assessment agents. Replace or complement written assessments with an oral conversation that can provide feedback and/or grades – with full educator controls, oversight, and professional judgement.
• Writing support agents. Oral agents that support students through key stages of writing: idea generation and pre-writing, planning and revision of drafted work. Students respond verbally to prompts and follow-up questions that help them develop and refine their thinking.

Scarlatti’s in-house oral AI agent, Vocina Views, delivers customised survey and interview questions verbally and records responses in a transcript-ready format for analysis. Qualitative, open-ended insights are key for organisations to make good decisions but collecting them can be slow, expensive and hard to scale. Vocina Views supports deeper insights at “survey scale” through natural conversation, follow-up questions, and consistent outputs that are ready for qualitative or mixed-methods analysis.

Many organisations see the potential of AI, but the hardest part is turning that potential into something practical, safe, and fit-for-purpose. Scarlatti provides hands-on support to help you identify the right use cases, design a solution that works in your environment, and implement it with confidence.

We can help with:

• Use-case discovery and prioritisation. Clarify where AI will genuinely add value (and where it won’t), with a focus on measurable outcomes.
• Agent design and development. Design and build tailored AI agents for specific workflows
• Prototyping and proof-of-value. Rapid prototypes to test feasibility, usability, and value before investing in full implementation
• Governance, safety, and quality. Support for evaluation approaches, human oversight, and practical risk management aligned to your context.