Language
Javaalura.com.br
Main backend language.
UI library
Reactalura.com.br
Front-end library.
Hosting
Cloud where the whole platform runs.
Kubernetesaws.amazon.com
Runs the backend microservices on Amazon EKS.
Terraformjob-boards.eu.greenhouse.io
Infrastructure as code for the AWS environments.
Database
MySQLalura.com.br
Relational database in the Java architecture.
Amazon DynamoDBjob-boards.eu.greenhouse.io
NoSQL database, alongside RDS and DocumentDB.
Backend framework
Spring Bootalura.com.br
Framework for the Java services.
File storage
Amazon S3job-boards.eu.greenhouse.io
File storage, alongside the CloudFront CDN.
Search
Amazon OpenSearchaws.amazon.com
Stores creators' lesson transcripts and feeds them to Bedrock.
AI model
Amazon Bedrockaws.amazon.com
Runs Claude Sonnet to answer students and route agents in Tutor.
Why: Comparison tests showed better results and lower infrastructure cost than the other options.
Error monitoring
Amazon CloudWatchaws.amazon.com
Monitors the Tutor infrastructure.
New Relicjob-boards.eu.greenhouse.io
Application monitoring.
Grafanajob-boards.eu.greenhouse.io
Dashboards on top of Prometheus metrics.
Hotmart has been an AWS customer since it was founded. Its backend services run as microservices on Amazon EKS, the managed Kubernetes from AWS.
A current DevOps opening lists the AWS pieces the team operates: RDS, ElastiCache, Elasticsearch, DocumentDB, DynamoDB, S3, IAM, CloudWatch and CloudFront, managed with Terraform and Docker.
In a recorded case for Alura, Marco Aurélio, a development specialist at Hotmart, explains how the company's Java architecture evolved along with the language, and the reasons behind each new technology.
The technologies covered in the talk include Java, JBoss, Spring Boot, Micronaut, MySQL, Docker, React and AWS.
Sources:alura.com.br
Hotmart built Tutor with AWS starting in December 2023 and launched it to Hotmart Club creators in August 2024. Lesson video transcripts are loaded into Amazon OpenSearch, and Amazon Bedrock with Claude Sonnet writes the answers and routes questions to specialized agents.
They chose Bedrock after comparison tests showed its models gave better results at a better infrastructure cost. Hotmart reports more than 17,000 creators turned it on, 1.6 million questions answered and up to 50% fewer support requests.
Sources:aws.amazon.com
Monitoring relies on New Relic, Prometheus and Grafana, and the security team is bringing AI agents into its NOC and SOC work.
Hotmart is also rebuilding how product teams work: it is hiring engineers whose main tool is a coding agent like Claude Code, Cursor or Codex, and who maintain the skills, templates and checks that turn business specs into production code.
Sources:job-boards.eu.greenhouse.io job-boards.eu.greenhouse.io job-boards.eu.greenhouse.io
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.
