OtherU Documentation

Welcome to the OtherU knowledge base. Discover guides, architecture overviews, and operational runbooks for running sovereign AI systems.

OtherU Docs Home

Return to the central hub and overview of the OtherU documentation ecosystem.

API and Integrations

Developer documentation for extending capabilities with APIs and third-party tools.

Architecture Overview

Deep dive into how system components interact securely and autonomously.

Connect Hermes Agent

Connect Hermes Agent to the inference stack and your messaging channels.

DeepSeek-V4 Quantization Pipeline

How OtherU quantizes DeepSeek-V4-Flash to its own affine fp2 format for Ember.

Ember — Inference Engine

OtherU's own C/C++ inference engine for DeepSeek V4 on AMD Strix Halo.

Answers to the most common questions about the OtherU ecosystem.

Short definitions for the terms used across these docs.

Hardware: Framework Desktop + jetKVM

Reference architectures and tooling for deploying on physical secure hardware.

The agent layer: channel routing, typed tools, and layered memory.

Hindsight Memory

Persistent long-term memory for Hermes Agent, powered by Hindsight.

Licensing terms, permitted use, and the disclosure policy.

Operations Runbook

Standard operating procedures, emergency protocols, and maintenance operations.

Products and Status

Current feature availability and system operational status updates.

Understand our approach to data privacy, network isolation, and local execution.

Setup and Deployment

Step-by-step guides for installing and initializing your local AI infrastructure.