Proca - Progressive Campaigning backend

An universal action tool backend for JAM stack apps. Elixir app that uses PostgreSQL as data store and RabbitMQ for data processing. It is the backend / datastore for Proca Widget.

It is made with love, elixir and hope we can change the world for the better.

Please note that this project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.

Contributor Covenant


  • Headless
  • GraphQL API
  • Fine grained permission system for users organised in Organisatons (Orgs)
  • Authentication using HTTP Basic Auth or JWT (to use external identity and auth system).
  • Stores campaign tree (action pages organised in campaigns, where different Orgs can run a coalition)
  • Stores actions and member personal data, personal data is E2E encrypted at rest. Only the Org that is data controller can decrypt it.
  • Validates personal data types using various personal data schemas (email+postcode, European Citizen Initaitive, and so on)
  • Handles GDPR consent (with optional double opt-in), and action staging (moderation, filtering before it is accepted)
  • Sends emails (thank you emails, opt in emails, emails-to-targets) through various mail backends (Mailjet, AWS SES, more coming)
  • Pluggable action processing on RabbitMQ queue
  • Forward actions to AWS SQS, CRMs (needs a decrypting gateway proxy at Org premises)
  • Export action data in CSV


  • PostgreSQL >= 10

  • Elixir >= 1.10 Erlang Solutions provides packages and binaries for download.

    You'll need the following packages: erlang-base erlang-dev erlang-parsetools erlang-xmerl elixir

  • RabbitMQ (3.x, tested on 3.8)

  • NodeJS (>= 10)

Just trying out setup


  • docker
  • docker-compose
  • proca-cli (install with: npm i -g @proca/cli)

If you would just like to try out proca server, it's easiest with docker-compose:

$ cd proca/utils
$ docker-compose up -d 
# wait until the servers start 

$ docker-compose logs proca 
# note down the username and password for the user of primary "instance" organisation. 

$ export API_URL=http://localhost:4000
$ proca-cli setup 
# in interactive session, choose "set up authentication" and then input
# "instance", "" and password you noted down:


- Using current working directory: /home/marcin/Projects/proca-server/proca/utils
- There is not .env file - I will create it after asking You some questions

? What would you like to do? Set up authentication (no org set, user is not set up, password is not set)
? What is the short name of your org? instance
? What is your username (email)?
? password: [hidden]
Thanks! Fetching campaign list to check the credentials
? What would you like to do? Save current config to .env file and leave

Now you can:

  • Use proca-cli (in the directory where .env file was created) to talk to the server API.
  • You can also perform API calls directly using GraphQL in the GraphQL playground - it's great for exploring the API! Sign in at http://localhost:4000 to make authenticated API calls.
  • You can see the processing in action using RabbitMQ management console - login with user name proca, password proca.

Development setup

The script utils/ will setup PostgreSQL, the Erlang / Elixir / Pheonix server, the RabbitMQ server and runs npm install in the assets directory.

The script needs a few ENV varaibles set:

$ export MIX_ENV=dev ORGANISATION='fix-the-status-quo'

You'll need sudo to run the parts of the script that configure PostgreSQL.

$ ./utils/

The seeds.exs command will print out your login and password:

#####   Created Admin user  #####
#####   Password: VERY_RANDOM_PASSWORD

You can then run the development server.

$ mix phx.server

By default, the development webserver is located at http://localhost:4000/


See config/config.exs and config/dev.exs for configuration options.