Education
Curriculum work spans teachers, schools, and districts. Teachers know how to teach. The school and district need the institutional memory of what's been built, the standards mapped to it, and the ability to pass an audit at the state or country level. Each year staff turn over and materials get rebuilt from scratch unless that work is owned somewhere durable.
We build a curriculum-authoring platform used in K-12 schools and districts. Teachers author through a mix of free-text drafting, structured forms, and prompt-based AI assists; the platform handles standards alignment, accessibility, and student-facing rendering automatically. The district owns the resulting library, so it survives staff turnover and passes audits at the state or country level. AI is part of how this works at scale: drafting from an outline, surfacing what other teachers in the district already built, and checking standards coverage before publishing. We start with one AI capability at a time (drafting, or alignment, or discovery), let it earn trust against actual classroom use, then expand; we don't roll the whole stack out at once.
Capabilities
Teachers shouldn't have to learn a content-management system to publish a lesson. The authoring experience mixes what each step actually needs: free-text drafting where prose is the point, structured forms where consistency matters, and prompt-based AI where a teacher's outline can be turned into a draft. Standards tagging, accessibility, and student-facing rendering all happen automatically underneath. The result feels like writing; the output is structured, audit-ready curriculum.
Lesson plans, rubrics, assessments, differentiated materials: repetitive to build from scratch, fast for AI to draft from a teacher's outline. The teacher refines and decides what fits their students. The drafting time goes to where pedagogical judgment actually matters: the parts the AI can't see.
When a district has hundreds of teachers building curriculum independently, the best work tends to stay on the shelf where it was created. AI-assisted discovery (semantic search, topic clustering, related-content suggestions) surfaces relevant materials a teacher didn't know existed. New teachers ramp faster; veteran teachers get reach for work they spent years building.
K-12 curriculum aligns to specific Common Core State Standards (CCSS) codes by ID and grade level, not to a category label. The platform tags each lesson, rubric, and assessment to the exact standard it covers and validates the alignment as the curriculum is authored. Coverage gaps surface across a unit, a course, or a whole grade level. The district stays audit-ready by default, not by year-end scramble.
The curriculum library exposes a Model Context Protocol (MCP) endpoint, so a teacher, principal, or district administrator can connect their own AI client (Claude Desktop, ChatGPT with MCP, an in-house assistant) and ask the library questions directly. “Which fifth-grade units cover this standard?” or “Which units mention fractions but don’t have an assessment yet?” get answered from inside the AI client they already use, instead of learning a new interface for each question.