TinyHumans

OpenHuman is a personal super intelligent AI with an inbuilt orchestration engine to manage all your agents in one place. It runs on a local machine and integrates with common apps to automate work flows, memory surfaces, and tool access. The tool emphasizes being open source and private, suitable for users who want to run on their own hardware.

OpenHuman connects to apps and services through a memory-driven architecture, enabling agents to use memory, context, and workflows to perform tasks. It provides a local-first approach where model calls can be directed to chosen providers, including locally hosted options, with a focus on privacy and control.

The project describes a memory engine that decides what to keep, how to compress it, and what surfaces when an agent needs context. It also includes a workflows engine that orchestrates agents to automate requests, with optional hosted or local deployment and integration with existing tools.

Office hardware and software environments are supported with downloadable builds detected by the page for macOS, Windows and Linux, and it highlights deep integrations and the ability to run a personal AI locally.

Use cases

• Automate personal and professional tasks across multiple apps using a single personal AI agent.