Stavros – Kubernetes, AWS, Terraform

Business travelers and people relocating for work need fully furnished apartments for stays of a month or longer, but traditional leases are rigid and hotels are expensive. The platform lets guests find, book, and manage furnished, move-in-ready apartments in major cities worldwide for flexible terms. It runs on cloud-native infrastructure on AWS and Kubernetes that serves guests and internal operations teams globally.

• Drove the adoption of a FinOps framework for cloud cost tracking and monitoring across the platform.
• Designed and implemented the migration of Lambda functions from Serverless Framework to Terraform, including automated deployments.
• Automated base image updates across the organization and introduced Renovate for automated dependency management across services.
• Added a network overlay to the Kubernetes platform.
• Established the company’s incident management policy.
• Contributed a cgroup-aware memory fix to an open-source search engine to improve its reliability in containerized environments.

An open-source team collaboration and messaging platform for organizations with strict security and compliance requirements, such as government, defense, and regulated enterprises. Many of these customers cannot use mainstream chat tools because they need control over where their data lives. The platform offers secure messaging, workflow automation, and integrations, delivered as a multi-tenant cloud SaaS on AWS across global regions as well as self-hosted.

• Architected compliant, enterprise-ready cloud infrastructure with a focus on data sovereignty, so customers with strict regulatory requirements could run on the SaaS offering.
• Led the migration of the multi-tenant cloud infrastructure from MySQL to PostgreSQL and optimized PostgreSQL performance with PgBouncer to increase cluster capacity.
• Led FinOps efforts that reduced cloud spend across global AWS workloads, and built a cost-monitoring stack with real-time budget dashboards.
• Built an SLO framework with reliability targets and incident-response metrics, and modernized disaster recovery with low-RTO strategies for critical services.
• Built and maintained FIPS-compliant infrastructure and hardened container images.
• Modernized CI/CD to speed up builds, migrated logging from Kibana to Loki, and tuned Kubernetes DNS policies to improve cluster performance and availability.

An online freelance marketplace where businesses find, hire, and pay independent professionals for projects in development, design, marketing, and other fields. It solves the slow and risky process of finding reliable freelancers by handling discovery, proposals, communication, billing, and payments in one place. The platform serves a large international base of clients and freelancers and had to absorb sharp traffic spikes during the pandemic.

• Migrated the core marketplace services to AWS, containerized the applications, and built CI/CD pipelines, then prototyped the first Kubernetes clusters and supported the move to microservices.
• Led platform optimizations that reduced cloud spend while handling the pandemic-era surge in traffic.
• Introduced SLIs, SLOs, and error budgets across services, plus disaster-day exercises to test recovery.
• Built the monitoring stack on Prometheus, Grafana, and Alertmanager, and hardened DNS configuration for better uptime.
• Developed PHP and Python backend services and internal APIs for billing, user management, and notifications, refactoring legacy code and raising test coverage.
• Introduced AWS WAF and a Redshift cluster to the platform.