Open-source repositories

Both LydianAI projects are open-source and self-hosted. Pick a project below and follow the setup guide to get running.

Federated ML

Server/coordinator + workers + CLI client. Runs across mixed machines — modern GPUs, Pascal GPUs, and CPU-only. Built on FedAvg + FastAPI over Tailscale.

Open on GitHub Quickstart

Key folders: server/, worker/, client/, common/

Assurance-as-Code

Backend for managing compliance artifacts for automotive SW Systems. Handles artifact lifecycle, traceability, and release-gate evaluation for UNECE R155, R156, and ISO 29119. Connects to GitHub repositories via a GitHub App — every push is scanned, implementation items are linked to requirements automatically, and compliance status is posted to PR checks.

Open on GitHub Docs

Self-hosted, Docker Compose deployment included

Getting started with Federated ML

1) Bring machines on the same network

Use Tailscale so the coordinator and workers can talk over stable 100.x addresses.

Tailscale setup

2) Pick NEW or LEGACY GPU install path

Modern GPUs use a current PyTorch stack. Pascal GPUs (GTX 1080/1080 Ti) require pinned legacy wheels and often Python 3.10.

GPU setup guide

Run the full Quickstart

Getting started with Assurance-as-Code

1) Deploy the backend

The repo includes a Docker Compose file that brings up the API, database, background job queue, and web console together. Clone the repo and run docker compose up to get started.

Deployment guide

2) Create your first SW System project

Use the API or web console to create a project, set the applicable standards (ASIL level, R155/R156/ISO 29119), and start adding artifacts. Import from Sphinx-Needs or create items directly through the API.

Project setup guide