Automatiser les processus métier.

Logo Activepieces

Activepieces

Self-hosted Zapier alternative with 280+ integrations and an AI agent hub — unlimited flows, zero per-task billing.

4 Go (2 Go minimum) RAM 2 vCPU Port 8080 Disponible

Stack technique

DockerNode.jsPostgreSQLRedis
RAM minimum4 Go (2 Go minimum)
CPU minimum2 vCPU
Port par défaut8080
OS compatiblesubuntu-24.04

Activepieces is an MIT-licensed workflow automation platform that replaces Zapier and Make on your own infrastructure. With 280+ integrations (pieces), a visual flow builder, and built-in support for AI agents and 400+ MCP servers, it covers everything from simple webhook-to-Slack pipelines to multi-step AI agentic workflows — all without usage-based billing.

Deployed on a ServOrbit VPS, Activepieces runs on a single Docker Compose stack (app + PostgreSQL + Redis). Your automation credentials, webhook payloads, and flow logic stay on your server. At a fixed monthly VPS cost, you can run thousands of automations that would cost hundreds of dollars on cloud platforms.

Fonctionnalités clés

280+ connectors (pieces) — Slack, Gmail, HubSpot, Notion, WhatsApp, MySQL, HTTP…
Visual flow builder with branches, loops, filters, and custom code steps in TypeScript
AI agent automation: 400+ MCP servers wired to your flows as native steps
Webhook receiver with a stable URL on your domain — no ngrok tunnels needed
Sandboxed code execution — run custom TypeScript steps safely in isolation
Unlimited flows and executions at a fixed VPS cost — no per-task billing
MIT-licensed core, self-hosted data, version-pinned updates under your control

Quand utiliser cette solution ?

1

Agency automation hub

Run client automation flows on your VPS instead of paying per-task. Handle webhook ingestion, CRM syncs, and WhatsApp notifications for every client at a flat monthly cost.

2

AI agentic workflows

Wire MCP servers (browser automation, code runners, web search) into your flows. Build reliable multi-step AI agents without writing a dedicated backend.

3

Replacing Zapier / Make

Migrate your existing automations flow by flow. Each of the 280+ pieces maps directly to a Zapier app or Make module — no rebuild from scratch.

Déployer Activepieces sur votre VPS

Guide optimisé pour les VPS Cloud ServOrbit.

01

Create the VPS

Order a ServOrbit VPS with at least 2 vCPU and 4 GB RAM. Ubuntu 22.04 LTS is the recommended OS. You will need a domain name (e.g. flows.yourdomain.ma) pointing to the VPS — Activepieces needs a public HTTPS URL to receive webhooks.

02

Deploy the stack

SSH into the VPS. Create /opt/activepieces/.env with your database credentials and two 32-character hex secrets (openssl rand -hex 16 twice for AP_ENCRYPTION_KEY and AP_JWT_SECRET). Run docker compose up -d — the app, PostgreSQL, and Redis start together.

03

Set AP_FRONTEND_URL (critical)

Add AP_FRONTEND_URL=https://flows.yourdomain.ma to your .env. This URL is embedded in every webhook Activepieces generates — if it's wrong, incoming webhooks will miss your instance. Restart after any change to this variable.

04

Put HTTPS in front

Install Caddy: apt install -y caddy. Create /etc/caddy/Caddyfile with: flows.yourdomain.ma { reverse_proxy localhost:8080 }. Caddy fetches a Let's Encrypt certificate automatically. Activepieces' webhook URLs will now be publicly reachable.

05

Create your first flow

Open https://flows.yourdomain.ma, create your admin account, then build a flow. Start with the Webhook trigger and connect a Slack or email step to validate your public URL. Check Settings → Connections to add your first integration credentials.

Questions fréquentes

Activepieces is an open-source (MIT) workflow automation platform. It lets you connect 280+ apps with a visual flow builder — similar to Zapier or Make, but self-hosted on your own VPS with no per-execution billing.

Automatiser les processus métier.

Activez Activepieces sur votre infrastructure.

VPS Cloud dédié — IPv4 incluse, datacenter européen, support inclus. Vos données ne quittent jamais votre serveur.

Configuration recommandée : 4 Go (2 Go minimum) RAM · 2 vCPU

Besoin d'aide ?

Parcourez notre centre d'aide et notre FAQ, ou écrivez à notre équipe — support en anglais, français et arabe.