IoT Device
Management Agent
Blazing fast and incredibly compact
The Ioto agent is the most powerful, intelligent device agent for device-based management with optional remote connectivity. Ioto includes an embedded HTTP web server, MQTT client, HTTP client, WebSockets, SSE, embedded database, JSON parsing, local, remote and hybrid AI, easy provisioning and OTA upgrading. All in under 300K code.
Blazing fast, compact and secure device agent
The Ioto Device Agent
Device agent HTTP and MQTT for local and remote management.
- Integrated embedded web server for direct local management.
- Optional remote management via MQTT.
- Transparent device data synchronization.
- HTTP Client for accessing external services.
- Embedded database for efficient state storage.
- WebSockets for interactive client-server communications.
- Server-Sent Events for push notifications.
- Simplified device and fleet provisioning.
- Blazing fast and compact (less than 300K code).
Embedded Web Server
For Local Device Management
- Fast, secure and compact embedded web server.
- HTTP/1.1 and SSL/TLS.
- WebSockets for interactive communications.
- Server-Sent Events for push notifications.
- Sessions and cookie management.
- Authentication and user management.
- High performance embedded database for device data.
- URL actions bind URLs to C functions with exceptional performance.
- Transfer chunking and file upload.
- HTML helpers for encoding, decoding and escaping.
- Tiny footprint: Web server component is 25K code.
- Unrivaled performance: >9,300 requests per second on a Raspberry PI.
- Easy extensions via the simple integration API.
- Hardened security due to extensive peer-review and fuzz stress testing.
The Embedded Database
High performance document database.
- JSON document items with flexible query API.
- Fast red/black binary search indexes.
- Database schema to define and validate data items.
- Controllable journal-based persistency and change triggers.
- Efficient import and export of database items.
- Automatic removal of expired data items.
- Per-table caching and buffering to control persistency.
- Simple, non-blocking API.
The HTTP Client
Connect to external services and data sources.
- Modelled on Browser Fetch API.
- Full support for all HTTP method verbs.
- WebSockets for interactive communications.
- Server-Sent Events for push notifications.
- Stream request and response body for large data transfers.
- Parallelism via fiber coroutines. No ugly callbacks or complex threads.
- Connect to LLMs for AI assisted services.
- Easily set request headers and access response headers.
- Support for JSON request and response bodies.
- SSL/TLS encryption.
- Tiny footprint: HTTP client component is 6K.
The MQTT Client
Connect to remote services to send or receive device data.
- Publish and subscribe MQTT messaging client.
- MQTT 3.1.1 compliant.
- SSL/TLS encryption with ALPN over port 443.
- Supports connect, publish, subscribe, ping and disconnect messages.
- Message quality of service for reliable delivery.
- Parallelism via fiber coroutines. No ugly callbacks or complex threads.
- Retained message support.
- Lower device connection cost via on-demand and scheduled connections.
- Elegant, non-blocking procedural API.
- Tiny footprint: MQTT component is 8K.
The Safe Runtime
Portable, cross-platform O/S abstraction layer.
- Efficient embedded runtime to protect against security threats.
- Battle-hardened after extensive peer review and stress testing.
- Safe strings, printf, lists, hashing and buffer management.
- Event and timer scheduling.
- Fiber coroutines for non-blocking procedural programming model.
- JSON parser, renderer and query engine.
- Flexible logging support.
- Global memory error handler.
IoT AI
Run local models, remote models or hybrid of both.
- Invoke local models on the device.
- Invoke remote LLMs from the device.
- Run agentic AI workflows on the device with agents and tools.
- Call Ioto device update, reboot, and other custom operations from LLMs.
- Implement hybrid AI with local inference escalating to remote services for powerful analysis and actions.
- Trigger remote LLMs based upon device data and metrics.
Fiber Coroutines
Parallelism without complex threads or callbacks.
- Support multiple requests with straight-line, non-blocking code patterns.
- Ioto services are fiber aware and automatically yield while waiting for I/O.
- Extremely fast fiber switching operations.
- Sandbox limits on resource usage.
- Low level fiber API to explicitly yield, resume and spawn fibers.
Hardware Support
Extensive micro-controller and embedded system support.
- Operating environments: ESP32, FreeBSD, FreeRTOS, Linux, Mac OS X, QNX, Quadros, VxWorks, Windows.
- CPU architectures: arm, arm64, mips, mips64, ppc, ppc64, riscv, riscv64, tidsp, sh, x86, x64 and xtensa.
- Extensive list of supported micro-controllers and CPU boards.
AI-First Development
AI as first-class users — quickly embed and extend
- Full AI context: CLAUDE.md, design docs, API doc with every agent.
- AI skills for security, testing, and coding.
- Rapidly create custom device management UIs with compelling UX.
App Scaffold
Get started fast with a pre-created app template.
The app scaffold includes everything you need to start building:
- Pre-installed Ioto agent release ready to build and run.
- Full AI context for all APIs and data structures.
- 17 AI skills: code audit, cross-compile, database, debugging, device UI, HTTP, JSON, MQTT, OpenAI integration, REST APIs, runtime, TLS configuration, UI spec, unit testing, WebSockets, and more.
- Makefiles and project build files for immediate development.
Ioto Agent Documentation
Read about
Get Started with the Ioto
Run stand-alone or with optional remote connectivity.
- Includes complete source code, documentation and samples.
- Free evaluation via the EmbedThis Builder.