Knowledge Architecture
ThinkBook
A structured knowledge substrate for curricula, concepts, prerequisites and pedagogical relationships — the canonical record of what education knows.
✦Architecture
The aime Operating System
A coherent stack — from teacher-facing applications down to the runtime substrate — designed to operate as a single educational intelligence system.
fig.01 — aime Operating Systemstack ↓
- L00 · SurfaceApplicationsaime Teachaime JAMaime Intelligence
- L01 · CoreOrchestrationLoom — Workflow Orchestration
- L02 · CoreAgentsKern — Agent Framework
- L03 · CoreReasoningLeo — Education ReasoningThinkCache — Reasoning Cache
- L04 · CoreKnowledgeThinkBook — Knowledge Architecture
- L05 · CorePedagogyEduRule — Pedagogical Engine
- L06 · SubstrateInfrastructureaime HubCloud ServicesEdge & Offline Services
↑ User-facing surfaceIntelligence coreRuntime substrate ↓
Intelligence Stack
The components
Each component is a discrete system with its own surface, contract and lifecycle — composed into a single platform.
Pedagogical Engine
EduRule
A rule and policy engine encoding pedagogical method, learning design and assessment logic that governs how the system teaches.
Reasoning Models
Leo
Education-specific reasoning models tuned for curriculum, classroom context and the constraints of teaching at scale.
Reasoning Cache
ThinkCache
An acceleration layer for reasoning workloads — reducing latency and cost across high-volume educational inference.
Agent Framework
Kern
The framework for composing educational agents: tutoring, lesson planning, assessment, intervention and operational workflows.
Orchestration
Loom
Workflow orchestration that binds agents, models, knowledge and data into reliable, observable educational processes.