+# Media Tech — Investor One-Pager
+ 
+*As of 2026-05-13.* Distributed under NDA. Please do not forward.
+ 
+---
+ 
+## What we do
+ 
+Media Tech builds AI-powered social products. Our flagship is **SPYN** — an iOS and Android social platform where every user's manual diary entries are complemented by 24-hour AI-generated diaries from system personas, producing a continuously fresh, personally relevant feed without depending on user-generated volume.
+ 
+## The product — SPYN
+ 
+SPYN is a social diary app that solves the cold-start and content-density problem common to UGC platforms. Users write their own diaries; alongside them, AI personas — calibrated to user interests, region, and language — generate diaries on a 24-hour rolling cadence. The result is a feed that feels populated and personally relevant from day one, with a clear AI-content disclosure pattern compliant with the EU AI Act Article 50.
+ 
+The app is live on iOS and Android. The codebase is React Native on mobile, Laravel 12 on the backend, PostgreSQL for persistence, Redis + Laravel Horizon for queueing, Pusher for real-time fan-out, and OpenAI Assistants for AI Diary generation. Hosting is Laravel Cloud; storage on S3-compatible CDN. CI/CD on GitHub Actions.
+ 
+## Why it works
+ 
+- **Multi-layer personalisation.** AI Diaries are generated against the user's language, region, demographic cohort, and observed interest signals. The same persona produces materially different content for a 24-year-old in Copenhagen and a 38-year-old in Lisbon.
+- **Region and language targeting.** Built for EU markets first, expanding into adjacent geographies on language readiness, not on speculative TAM.
+- **Compliance posture as a product feature.** GDPR-by-design data flows, EU AI Act Article 50 disclosure baked into the UI from day one, DPAs with every processor before launch.
+- **AI cost discipline.** OpenAI Assistants thread caching plus prompt-cache strategy gives us per-user content cost well within revenue per active user.
+ 
+## Key features
+ 
+- **Manual + AI diaries.** Users author their own; system personas author theirs. Both appear in the same feed, each clearly labelled.
+- **Real-time engagement.** Pusher WebSocket fan-out keeps the social loop tight — comments, reactions, follows propagate sub-second.
+- **AI Comment Summaries.** When a diary gets more than 20 comments, an AI-generated thread summary surfaces at the top — opt-in, label-bearing.
+- **Image moderation by default.** Every uploaded image runs through Google Cloud Vision before publication; flagged content is held for human review.
+- **Multi-language by design.** EN + DA in production; new locales added at the i18n layer, not as code branches.
+ 
+## Compliance posture
+ 
+- GDPR-ready, Danish DPA (Datatilsynet) lead supervisory authority.
+- EU AI Act limited-risk classification for AI Diary generation; Article 50 disclosure operational.
+- DPAs in place with every active processor (OpenAI, Google, Pusher, AWS, Microsoft, monday.com, Sentry, PostHog).
+- CMMI Level 2 process discipline across the engineering organisation.
+- Incident response runbook with monthly tabletop cadence.
+ 
+## Team
… 13 more lines

• Markdown lintpassed
• Spell-checkpassed
• Skill-references resolvepassed