Cauri Money runs a transactional platform where downtime is not an option. The legacy platform was deployed manually: every release was a source of errors and stress, and the VMs were heavily oversized as a safety margin — needlessly inflating the cloud bill.
We designed and deployed a multi-AZ AWS architecture with Load Balancers, auto-scaled EC2 nodegroups, S3 and CodeDeploy, together with a full CI/CD pipeline (build, tests, automated delivery to each environment). Auto-scaling and right-sizing brought consumption back in line with real usage, significantly reducing infrastructure costs; the pipeline removed the stress and errors of manual production releases.
Azure DevOps anchored the whole team workflow: repositories, access and rights management, GitFlow (branches, pull requests, branch protections), mandatory code reviews and CI/CD pipelines. A clear, shared frame to ship with confidence.
For workloads with high traffic variability, we designed a 100% serverless architecture. CloudFront at the edge, API Gateway for routing, Lambda for compute.
In-flight engagement: design and industrialization of a unified data platform.
The platform is built on a proven open-source stack: Airflow, DBT, and a dual Postgres / ClickHouse foundation.
kubeadm or rke2 install, networking, storage, backup and DR.
Multi-stage Jenkins or GitLab CI pipelines, GitOps with ArgoCD.
Claude adoption for product, support and ops teams.
Prometheus + Grafana + Loki stack, business dashboards.
EKS hardening, network policies, secrets management.
Lift & shift, AWS replatform, progressive refactor.
Every serious Cloud / DevOps project is interesting to us. Let's talk.
Discuss the project