Language
Main backend language.
Pythonjob-boards.greenhouse.io
Backend of the AI agent builder and other AI services.
UI library
Front end of the products, with TypeScript.
Hosting
Google Cloudjob-boards.greenhouse.io
Cloud for the products, alongside AWS and Oracle Cloud.
Kubernetesjob-boards.greenhouse.io
Container platform for many products, on GKE and EKS.
Terraformjob-boards.greenhouse.io
Infrastructure as code with reusable modules for engineering teams.
Database
PostgreSQLjob-boards.greenhouse.io
Main relational database.
MongoDBpt.slideshare.net
Document database, which held a billion documents back in 2015.
Google Cloud Bigtablegithub.com
High-volume database that RD autoscales with its own open source Kubernetes operator.
Data fetching
Sidekiqjob-boards.greenhouse.io
Async job processing for the Rails apps.
Event streaming between services, alongside Google Pub/Sub.
Backend framework
Ruby on Railsjob-boards.greenhouse.io
Framework of the core products, including the billing system.
Error monitoring
Datadogjob-boards.greenhouse.io
Monitoring and performance in production.
Rollbarjob-boards.greenhouse.io
Error tracking in production.
At TDC 2014, RD engineers showed how the stack had grown. In early 2013 it was just Postgres, Rails, Heroku and Delayed::Job. By March 2014 it had added Elasticsearch, Redis, MongoDB, memcached, Resque, GitHub and CircleCI.
The lessons on the slides point the same way: separate apps, decouple the front end from the back end and automate deploys.
Sources:speakerdeck.com
By 2015 RD Station was sending more than 70 million emails a month, running over 4 million jobs a day and holding a billion documents in MongoDB, with five dev teams deploying about ten times a day.
To keep up they leaned on autoscalers and microservices, trying Ruby first, then Node.js, then Go. New Relic profiled the app, Rollbar tracked errors and Hubot handled deploys.
Sources:pt.slideshare.net
TOTVS bought RD Station in 2021 for R$ 1.86 billion. Rails is still at the core: a current billing role asks for Ruby on Rails with PostgreSQL, Sidekiq for async jobs, Docker and Kubernetes, with Datadog and Rollbar in production.
The cloud team runs a shared foundation across Google Cloud, AWS and Oracle Cloud, with Kubernetes on GKE and EKS and Terraform for infrastructure as code. RD also open-sourced a Kubernetes operator that autoscales Google Cloud Bigtable nodes based on CPU.
Sources:olhardigital.com.br job-boards.greenhouse.io job-boards.greenhouse.io github.com
AI work happens in a cross-product group called the AI Alliance, which builds RD Station's AI agent builder. Its backend is in Python with React and TypeScript on the front, Kafka or Pub/Sub for events, PostgreSQL and MongoDB for data and Datadog for observability.
Engineers are expected to use AI tools like Cursor to move faster while staying responsible for quality and security.
Found on the homepage with the open Wappalyzer rules (HTML, scripts and headers). It shows what runs in the browser, not the whole backend.
Checked on 2026-10-29. Every tool links to the public source that shows it.
