Live 2025

Interlinear

Learn Latin From Caesar. Eventually.

The language tutor that knows why you're wrong—not just that you are

5 error types

Different fixes

typo ≠ case error ≠ vocabulary gap

Any text

Becomes a course

exercises, dialogs, vocabulary

Classical

Languages that work

Latin morphology, not pattern matching

Ever wanted to learn Latin from Caesar? Practice Greek with Cleopatra? Have Cicero correct your subjunctive?

That's where this is going.

But first, the foundation. Every language app gives you the same feedback: ❌ Wrong. Try again. That's not teaching—it's a slot machine with educational branding. The difference between a typo and a conceptual gap is the difference between "you fat-fingered it" and "you don't understand accusative case." A human tutor sees this instantly. Duolingo doesn't try.

Interlinear does. When you write puellam instead of puella, it identifies a case error and explains why nominative was expected. Different errors, different interventions. The same evaluation harness patterns we use for production LLMs—applied to language learning.

Then: course generation. Upload any reading and the system generates a full course—comprehension exercises, translation drills, contextual dialogs, vocabulary cards. All calibrated to your demonstrated level.

Then: the dialogs come alive. AI-generated talking heads. Real-time conversation practice with historical figures, native speakers, patient tutors who never get frustrated. The error correction, the course generation, the pedagogical engine—all of it feeds into conversations that actually teach.

First course in development: Introduction to Old Norse. Academic rigor. Research preparation. And eventually, a conversation with a skald.

Architecture

Mastra orchestrates multi-stage content generation and error analysis. The error classifier distinguishes five categories—each triggering different pedagogical responses. CEFR calibration is continuous and per-skill (you might be B2 vocabulary but A2 subjunctive). Classical language support includes real inflection analysis, not pattern matching.

View diagram
Interlinear architecture diagram

Key Decisions

Error Taxonomy Over Binary Feedback

Production LLM evals distinguish retrieval failures from reasoning errors from formatting issues. Same logic applies to language learning. Different root causes need different interventions.

Course Generation From Any Text

Fixed curricula are the bottleneck. Let students learn from texts they actually care about. The platform becomes the course authoring tool.

Classical Language First

Latin morphology is the hardest test case. If you can handle 50+ forms per noun and proper case analysis, modern languages are trivial. Built the hard thing first.

Continuous CEFR Calibration

Students are uneven. A2 in one skill, B2 in another. The system tracks competency per grammatical concept, not per student level.

What Was Hard

Generating coherent course progressions, not just random exercises. Each generated course needs narrative structure—skills build on each other. Still iterating on progression algorithms.

Claude Sonnet Vercel AI SDK Mastra Workflows ElevenLabs TTS Supabase Next.js 15 Cloud Run