Delivered projects

A multi-tenant SaaS platform built from scratch on self-hosted open-source infrastructure

One continuous engagement: architecture, backend, frontend, infrastructure and SRE. The platform is in production and still growing, and the team stays embedded long-term, shipping new features continuously.

• Microservices: 3 backend services in Go (delivery engine, billing/CRM, campaign management) and a Next.js/React dashboard plus public landing page, bilingual EN/RO.
• Strict tenant isolation, verified with dedicated per-route tests, and 600+ automated tests across all backend services.
• 100% open-source, self-hosted infrastructure with no vendor lock-in: HA Kubernetes with a 3-node control plane, PostgreSQL ×3, RabbitMQ, Keycloak. Automated DNS (SPF, DKIM, DMARC) with TSIG-signed dynamic updates.
• Production cluster on the RunCloud.AI platform (Timișoara): 9 active nodes, 48 vCPU and 217 GiB RAM allocatable, currently using ~3.5 vCPU and ~18.7 GiB, leaving ample headroom for growth. A dedicated namespace with ResourceQuota and LimitRange sized to ~50% of cluster capacity.
• Deployment with 3 replicas, zero-downtime rolling updates, readiness/liveness probes and pod anti-affinity for node-failure resilience. A HorizontalPodAutoscaler between 3 and 12 replicas at 70% CPU utilization.
• SRE from day one: SLO-based alerting (error-budget burn rate), an incident runbook built from real incidents, in-house Prometheus-compatible metrics, Loki and automatic log-based error classification.
• Self-hosted AI with Ollama for sales chat and reply classification, with zero external API cost.
• A native CRM integrated into billing, with no third-party CRM dependency, and complete technical documentation (C4 diagrams, ADRs).

• Go
• Next.js
• React
• Kubernetes
• CloudNativePG
• PostgreSQL
• RabbitMQ
• Keycloak
• Ollama
• Prometheus
• Loki
• HPA
• Docker Compose

Fully managed infrastructure from scratch for a B2C startup

A Kubernetes cluster on DigitalOcean built with Terraform, GitOps with Argo CD for staging and production, load balancing and Nginx Ingress, dev, stage and prod environments. A PostgreSQL cluster, ELK, RabbitMQ and S3, monitoring with Prometheus, Grafana and Zabbix, and CI/CD optimized on GitLab for 7 microservices.

• Kubernetes
• Terraform
• Argo CD
• Helm
• PostgreSQL
• ELK
• RabbitMQ
• Prometheus
• Grafana
• Zabbix
• GitLab CI/CD
• DigitalOcean

On-premise compute and Big Data platform

Development and operation of an on-premise Kubernetes compute platform, using Konvoy (D2iQ) to provision clusters as infrastructure as code. Ansible modules, GitLab CI/CD for the Big Data and analytics platform, TCP/IP networking, 9x5 support and 24/7 on-call.

• Kubernetes
• D2iQ Konvoy
• Ansible
• GitLab CI/CD
• Thanos
• Prometheus
• Grafana
• PostgreSQL

API Management and security on AWS and Azure

Infrastructure as code for API Management, Terraform pipelines with reusable building blocks for AWS and Azure, analysis of RESTful API traffic and network errors, plus resolving Prisma Cloud security alerts (Route 53, EKS, S3, IAM, EC2).

• AWS
• Azure
• Terraform
• Prisma Cloud
• Grafana
• Prometheus
• Splunk

Microservices fully managed on RunCloud.AI

A microservices architecture running entirely on the RunCloud.AI platform: dev, stage and prod environments, a PostgreSQL cluster, Dockerfiles for the Admin UI and backend API, a GitOps pipeline on GitLab CI/CD, 9x5 support with 24/7 on-call, and a post-implementation support subscription.

• RunCloud.AI
• Kubernetes
• PostgreSQL
• GitLab CI/CD
• GitOps
• Docker