Built
- Private Linux host with containerized services
- Home Assistant and MQTT integration
- Adaptive Next.js dashboard foundation
- LifeEvent data model
- Local and cloud AI routing
- Secure Tailscale network access
Functional prototype · Active development2026
RediON²⁶
Private intelligence and personal life operating system
Role
Product Architecture · Full-stack · AI · IoT · Infrastructure
01
Project overview
RediON²⁶ is a private intelligence system designed to bring smart-home control, personal AI, health, work, documents, media and connected devices into one coherent environment.
Instead of relying on disconnected cloud applications, the system runs on private infrastructure and exposes one adaptive Next.js interface across desktop monitors, televisions, tablets, phones and small ESP32 displays.
Its central LifeEvent model turns daily activity into structured, useful context. Local AI can discover patterns and produce summaries, while cloud AI is reserved for selected real-time tasks.
02
Real project media
＋
Real photographs and screenshots for this project will be added here.
03
System architecture
Sensors · ESP32 · Cameras · Piano
Physical inputs and devices
Home Assistant · MQTT
Automation and event transport
RediON²⁶ Core Server
LifeEvents · Local AI · Cloud AI router
Desktop · TV · Mobile · TFT
One adaptive interface
04
Project status
In active development
- Expanded health and routine analytics
- Document intelligence workflows
- PLUTO assistant orchestration
- Deeper device and piano integrations
Planned
- Proactive cross-module insights
- More custom ESP32 interfaces
- Kitchen and garden intelligence
- Advanced voice and vision automations
05
Engineering decisions
Why self-hosted?
Personal health, home and activity data remain under direct control while local services keep essential functions available without a public cloud dependency.
Why one LifeEvent model?
A shared event language allows health, environment, routines and devices to be analyzed together without tightly coupling every module.
Why combine local and cloud AI?
Local AI protects privacy and handles persistent analysis; cloud models are used selectively where stronger real-time capabilities add clear value.
06
Challenges solved
- 01Designing one interface for screens from 2.4 to 70 inches
- 02Keeping private services reachable without exposing public application ports
- 03Normalizing events from unrelated devices and modules
- 04Separating local intelligence from selected cloud operations
07
Technology stack
Application
- Next.js
- TypeScript
- React
- Prisma
- SQLite
AI
- Ollama
- OpenAI API
- Local vision
IoT
- Home Assistant
- MQTT
- ESPHome
- Frigate
Infrastructure
- Linux
- Docker
- Tailscale
08
What it demonstrates
RediON²⁶ demonstrates my ability to design a complete private computing environment across product architecture, AI, data modeling, infrastructure, IoT and adaptive interface development.