# Mentra

> Mentra is an AI-native, multi-tenant headless CMS designed for structured educational content. It empowers creators to build complex learning hierarchies powered by a suite of specialized AI agents.

## For AI Agents

- [llms-full.txt](https://www.mentra.systems/llms-full.txt) — complete Public API v1 documentation as a single plain-text document (recommended for LLM ingestion)
- [OpenAPI JSON](https://api.mentra.systems/api/v1/openapi.json) — machine-readable spec of every public endpoint
- [Changelog](https://api.mentra.systems/api/v1/changelog) — JSON feed of API changes

## Core Capabilities

- **Structured Content Hierarchy**: LifeSkills → Courses → Journeys → Steps — a 4-level depth model for educational content.
- **Headless CMS**: RESTful Public API for delivering content to any frontend, mobile app, or device.
- **Multi-Round AI Intelligence Engine**: Not a GPT wrapper. Every AI stage runs multiple adversarial rounds with constraint injection and ground-truth protection.
- **Strategic Research**: Persona analysis, competitive intelligence, market positioning — all AI-driven.
- **Content Drafting**: Template-based content generation calibrated to brand voice.
- **Multi-Modal Learning**: Automatic generation of mnemonics (text, image, video) and assessments (quizzes, real-life scenarios).
- **Multi-Tenancy**: Native multi-tenant architecture with Role-Based Access Control (RBAC) and tiered Stripe subscriptions.

## AI Pipeline Stages

1. **UNDERSTAND** — 3-round brand identity analysis: Extraction → Persona Deepening → Persona Sharpening. Produces Brand Profile, Voice DNA, Audience Personas, Visual Identity.
2. **CONTEXTUALIZE** — 4-round competitive intelligence: Collection → Qualification → Interrogation → Exploitation. Maps market landscape and competitive positioning.
3. **GOALS** — Campaign strategy definition: objectives, audience targeting, campaign visual themes.
4. **DECIDE** — 3-round strategy finalization: Hypothesis → Stress Test → Final Lock. Adversarial validation of strategic decisions.
5. **LEARN** — 2-round insight extraction: Explain → Prescribe. Turns pipeline outputs into actionable learnings.
6. **STRATEGY** — Orchestration layer coordinating all stages with constraint injection (Voice Calibration, Audience Intelligence, Campaign Intent).
7. **CONTENT** — Structured and standalone content generation calibrated to brand voice and hierarchy role.
8. **SOCIAL** — Platform-aware social content: repurposing, original posts, hooks — with per-platform character limits and rules.

## Key Differentiators

- **Adversarial Rounds**: Each AI stage uses multiple rounds that challenge and refine outputs, not single-pass generation.
- **Constraint Injection**: Brand voice, audience intelligence, and campaign intent are injected as constraints into every generation round.
- **Ground Truth Protection**: Human-approved data is never overwritten by AI — only extended.
- **Approval Gates**: Critical outputs require human approval before downstream stages can use them.

## Links

- [How Our AI Works](https://www.mentra.systems/how-our-ai-works) — detailed methodology page explaining each pipeline stage
- [Public API Documentation](https://www.mentra.systems/public-api-docs) — human-facing RESTful API reference (HTML)
- [Pricing](https://www.mentra.systems/#pricing) — subscription tiers and feature entitlements

## Contact

- Website: https://www.mentra.systems