The open-source stack
for AI agents.

What Supabase did for databases, neurwerk.base does for AI agents. Twelve integrated layers — frontends, triggers, proxy, LLMs, monitoring, memory, runtime, secrets, security, auth, and inter-agent communication — all fully open-source, self-hosted on your infra.

Built on battle-tested open-source tooling

✦ 12-layer modular architecture ✦ Fully open-source (not open-core) ✦ GDPR & EU AI Act compliant ✦ Self-hosted on your infrastructure

Easy deployment. With the CLI.

terminal — neurwerk.base quickstart

Every layer. Open and integrated.

neurwerk.base is a curated, open-source stack organised into twelve layers. Each layer solves a specific concern — from user-facing interfaces to agent runtime — using the best available open-source tools, pre-configured to work together out of the box.

01 – 03 Interfaces
01

Administration

Central admin UI for managing agents, services, memory, and team access.

Studio
02

Frontends

User-facing chat interfaces and interaction surfaces for agents.

LibreChat Dify n8n Gitea tba
03

Triggers

How agents are invoked: scheduled jobs, webhooks, events, and chat commands.

signal-cli Gitea tba
04 – 07 Core Engine
04

Proxy Layer

LLM and MCP routing, load balancing, and HTTP reverse proxy for all agent traffic.

AgentGateway K8S Gateway
05

Memory

Persistent memory on client, team, user, and agent level. Each level with its own prompt memory, session archives, skills, and external knowledge.

PostgreSQL Gitea
06

Runtime

Sandboxed agent execution environment with tool access and lifecycle management.

Goose Gitea
07

Inter-agent

Custom agentic email messaging system for inter-agent communication.

A-mail
08 Observation
08

Monitoring

Observability, token tracking, cost analysis, and alerting for every agent call.

Langfuse
09 – 12 Foundation
09

Secrets

Encrypted credential storage and secure injection into running agents.

SOPS age
10

Security

Threat model, network policies, and hardening controls across the stack, with layered PII protection.

AgentGateway Presidio
11

Auth & Users

User provisioning, SSO, roles, and access rights across the whole stack.

Keycloak
12

LLM Providers

Language model providers available to every agent: local, private, or frontier.

Llama.cpp OpenRouter Mistral OpenAI Anthrophic

Ship agentic apps like it's nothing

Every layer of the stack comes with production-grade capabilities, pre-integrated and ready to deploy.

🔐

Auth & User Management

SSO, OAuth, JWT, and role-based access for your entire agent fleet, frontends and custom MCP servers. Connect any identity provider. Built on Keycloak.

📊

Compliance & Security

Real-time token tracking, cost analysis, and usage monitoring per agent and per user. Combined with smart LLM routing across providers, content filters, PII redaction, and jailbreak detection at the proxy layer.

🤖

Agent Orchestration

Define, deploy, and connect agents across functions through neurwerk.studio. Use a wide variaty of triggers such as email or text to activate your agents.

🧠

4-Layer Memory

Client, team, user, and agent-level memory with prompt injection, full-text session archives, reusable skills, and PostgreSQL/pgvector external storage.

🌍

Host is your way

Deploy on AWS, GCP, Azure, or bare metal. No vendor lock-in. Full data sovereignty. EU-first by design.

🔌

Provider Agnostic

Ollama, OpenAI, Anthropic, Mistral, Llama or Gemini, one unified interface. Switch models with a config change.

Ready to build?

Deploy the full stack today. Self-hosted, open-source, and production-ready.

Star on GitHub →

Forever fully open-source.