___              __   __              _____         _
 / _ \__      ___ _\ \ / /__  _   _ _ _|_   _|__  ___| |__
| | | \ \ /\ / / '_ \ V / _ \| | | | '__|| |/ _ \/ __| '_ \
| |_| |\ V  V /| | | | | (_) | |_| | |   | |  __/ (__| | | |
 \___/  \_/\_/ |_| |_|_|\___/ \__,_|_|   |_|\___|\___|_| |_|

› your data. your ai. your productivity.

OwnYourTech is an MCP server that plugs into Claude Code, Cursor, Codex — or whatever AI workflow you already use. Ask it to build a data pipeline and it does: dlt, dbt, DuckDB, ready to run.

Based on real human experience — put into workflows, recipes, and best practices. Everything on infrastructure you control. Open-source tools, no vendor lock-in.

ls use-cases/

01-data-pipelines Your Data 
> "build me an analytics pipeline from our CRM"

→ Using OwnYourTech MCP...
→ Source: CRM REST API (token auth, cursor pagination)
→ Storage: data/analytics.duckdb

  Generated files:
  extract/pipeline.py
  transform/models/staging/stg_contacts.sql
  transform/models/staging/stg_deals.sql
  transform/models/analytics/rpt_sales_pipeline.sql
  validate/checks.sql
  run.sh

[x] Pipeline ready. Run: ./run.sh
02-local-ai Your AI — coming soon 
> "set up a local LLM for our support docs"

→ Using OwnYourTech MCP...
→ Model: Llama 3 via Ollama on your machine
→ Embedding: your docs into local vector store

  Generated files:
  ollama/modelfile
  embeddings/index_docs.py
  chat/app.py
  docker-compose.yml

[x] Local AI ready. No data leaves your network.
03-email-server Your Productivity — coming soon 
> "set up my own email server on Hetzner"

→ Using OwnYourTech MCP...
→ Provider: Hetzner Cloud (Nuremberg)
→ Stack: Stalwart Mail Server + Caddy

  Generated files:
  infra/docker-compose.yml
  infra/caddy/Caddyfile
  infra/stalwart/config.toml
  setup.sh

[x] Email server ready. Your inbox, your rules.

› Works inside Claude Code, Cursor, or Codex. Every generated file is plain code on your machine — nothing hidden.

cat how-it-works.txt

WHAT IS IT
----------
OwnYourTech is an MCP server. It plugs into Claude Code, Cursor,
Codex — or whatever AI workflow you already use.

It's packed with expert knowledge from real practitioners:
workflows, recipes, and best practices for setting up your
own infrastructure. No guessing, no hallucinating configs.

YOUR DATA
---------
Build data pipelines with proven open-source tools:
→ dlt    — extract data from any API or file
→ dbt    — transform and test with SQL
→ DuckDB — fast local analytics, zero infrastructure

YOUR AI (coming soon)
---------------------
Run your own models, on your own hardware:
→ Ollama — local LLMs without cloud dependencies
→ Vector stores — your docs, your embeddings
→ No data leaves your network

YOUR PRODUCTIVITY (coming soon)
-------------------------------
Own the tools you use every day:
→ Email — Stalwart on Hetzner, not Gmail
→ Hosting — your sites on European clouds
→ No subscriptions you don't control

WHY SOVEREIGN
-------------
Everything runs on infrastructure you choose — Hetzner,
Scaleway, OVH. Switch providers without rewriting your stack.

→ Open source: MIT licensed
→ VC-free: building trust, not exits
→ Foundation + company model (like DuckDB)

who --contributors

FOUNDERS
--------
@julienhurault Data Engineering, OwnYourTech
@timo-dechau Growth, Product, Event Data Modeling and Analytics 
WANT TO JOIN?
-------------
We're looking for:
  • → Data engineers with production experience
  • → European cloud expertise (Hetzner, Scaleway, OVH)
  • → Governance & compliance specialists

$ apply Contact us

$ apply --contributor GitHub