v0.1.0 · local-first IoT

Project MagNET

An open, local-first platform for IoT data and hyperlocal, hive-centric AI. Discover, provision, control, and sync your devices — without proprietary clouds, and on your terms for data and security.

Open Source — Apache 2.0 Licensed.

What is Project MagNET?

A constellation of cooperating edge devices — phones, wearables, sensors, and microcontrollers — that organize themselves, share data locally, and run intelligence at the edge. Every device is a cell in a larger, hyperlocal nervous system.

cloud_off

Local-First

Your data stays where you are. MagNET captures and syncs IoT data on your own subnet — local or in a cloud you control — so you avoid lock-in to proprietary services.

hive

Hive-Centric AI

Devices self-organize into a hive under a ruler, hot-swap roles over the air, and share memory. Not every device needs to be a brain — each acts as a cell, neuron, or synapse in a swarm.

my_location

Hyperlocal Context

Awareness scoped to here and now — locations, sightings, and a transaction model for advertising and offloading compute and AI between nearby devices.

cloud

Every Device, a Cloud

Any device can serve a cloud of services, discoverable over BLE or Wi-Fi. Detect, connect, provision, control, and collect — the five things a MagNET node should always be able to do.

lock

You Own Your Data

Security on by default — secure transport and verified session-key exchange. It’s up to each node to decide how much it trusts a given data sink.

developer_board

Reference Designs

Working firmware for ESP32, ESP32-S3/-C3/-C6, and nRF52 boards — wearables, beacons, time servers, Thread/CoAP meshes, data sinks, and the Hive AI prototype.

head_mounted_device

WebXR of Things

A spatial, WebXR front-end for exploring and configuring hyperlocal data and devices — the future we were promised, for the IoT around you.

extension

Open & Interoperable

Built to integrate with open standards and ecosystems — zeroconf/Bonjour discovery, LoRa/Meshtastic, Thread, Zigbee, and generic serialization.