UnoMundi
We're building Una: a scalable, schema-driven learning system for a global children's learning platform. Una integrates narrative design, developmentally informed pedagogy, and AI-enabled content workflows to support personalized, mastery-based learning at scale, without relying on traditional tests.
Responsibilities
1
Architecture and Delivery
Own architecture and delivery of Una, integrating narrative design and developmentally informed pedagogy into a schema-driven content system.
2
System Governance
Build and govern AI-supported curriculum generation pipelines; prompt systems, content schemas, and human-in-the-loop validation, in partnership with engineering and product.
3
Skills Progression
Design skill-progression frameworks and formative learning signals that enable AI-assisted mastery tracking without traditional testing structures.
UnoMundi
Una is a schema-driven AI learning system I'm building for a global children's learning platform - pairing narrative design and developmentally informed pedagogy with an AI-supported content pipeline that can generate and govern learning content at a scale no manual team could sustain alone.
1
The Challenge
-
Traditional learning platforms lean on quizzes and formal assessments to measure progress. For young learners, that model creates friction: it interrupts the learning experience and doesn't reflect how children actually build skills over time.The challenge was to design a system that could track mastery developmentally, through how a child engages with content, rather than through pass/fail checkpoints, while still generating personalized content at a scale no manual content team could sustain alone.
2
The Approach
Una is built around a content schema that pairs narrative design with developmentally informed pedagogy, so that AI-generated content stays both engaging and age-appropriate.
-
Skill-progression frameworks: structured pathways that define what mastery looks like at each stage, independent of test scores.
-
Formative learning signals: in-context behavioral signals used to infer progress, feeding AI-assisted mastery tracking.
-
AI-supported curriculum generation pipeline: prompt systems and content schemas generate draft content at scale.
-
Human-in-the-loop validation: every AI-generated unit passes through a structured review step before it reaches a learner, aligned to product strategy and pedagogical standards.
3
Process
-
Framework design: Defined skill-progression structures and formative signals in partnership with pedagogy and content stakeholders.
-
Pipeline architecture: Built the prompt systems and content schemas that drive AI-supported curriculum generation.
-
Governance design: Established the human-in-the-loop validation step to catch pedagogical, narrative, or safety issues before content ships.
-
Cross-functional delivery: Partnered with engineering and product to keep the pipeline aligned to product strategy as requirements evolved.
4
Where It Stands
The pipeline currently tracks 5,533 minutes (~92 hours) of lesson content in active production. 73.9% (4,089 minutes / ~68 hours) is fully complete and signed off; the remainder is distributed across review and revision stages: gallery assets and video content awaiting revision, review, or final production.
Get in Touch
This is a Paragraph. Click on "Edit Text" or double click on the text box to start editing the content.