Project experience (continued)
Greenfield Rails application with business stakeholders
2025 - 2026
MTÜ Ruumiharidus · via Whyservices OÜ
Industry: education · Role: Rails developer, from discovery to operation · Team: solo, with two stakeholders on the business side
The non-profit ran its core process by hand in spreadsheets and email, with a WordPress plugin managing students, groups and timetables. Double entries and errors cost the team 16 hours a week, and nobody had a reliable overview of what the software could do. I worked out the requirements directly with the stakeholders: short sessions, written specifications and small releases they could try out early. I built the application as a conventional Rails monolith covering registration, enrolment, attendance and invoicing, in Estonian and English. I also set up hosting, CI and monitoring, so the client had one contact from the first conversation to operation. Live after 12 weeks, with 2 people using it daily. Manual work dropped by two thirds.
Back-office application for a hosting business
2015 - today
Whyservices OÜ, own product (WellPress / WellHost)
Industry: hosting & IT services · Role: product owner & sole developer · Team: solo
Running managed hosting for many small clients meant invoicing, client communication, status updates and each site's hosting history were spread across separate tools. Every new client added manual work and the risk of a missed invoice or an unannounced maintenance window. I designed and built a Rails back-office application that brings these together: invoicing, client communication, public status pages and a per-site hosting history. It calls the hosting providers' APIs and manages the administration runbooks. I kept it plain Rails, so that maintaining it stays cheap for a one-person business. One person runs 30 client sites and applications. Setting up a new project takes 30 minutes instead of 8 hours, and invoicing and status communication are automated.
Container hosting platform for small companies (WellHost)
2022 - today
Whyservices OÜ, own service · Tallinn, remote
Industry: hosting for SMEs and agencies · Role: architect & sole operator · Team: solo
Small companies ran their applications without operations staff: hand-built servers, manual deploys, no separate staging, and backups nobody had ever restored. A failed server or a bad deploy meant days of downtime and possibly lost data. I built a standardised container platform on Docker Swarm, with CapRover as management layer and servers set up by Ansible. Each client gets staging and production separated at infrastructure level, with backing services running outside the application containers. Deploys run from git through CI (Woodpecker, GitHub Actions, GitLab CI) with rollback. Hourly encrypted off-site backups are restored in regular drills, and TLS, monitoring and weekly updates are part of the service. At most one hour of data can be lost, and restores are proven rather than assumed. 10 applications (e.g. Rails, Symfony, Django, Wiki.js, GitLab, Plausible) run at 99.95% availability. Clients pay from €50 a month instead of a multiple for dedicated operations staff.
Managed WordPress fleet (WellPress)
2013 - today
Whyservices OÜ, own service · Tallinn, remote
Industry: agencies, SMEs, non-profits · Role: architect & sole operator · Team: solo
Agencies and small businesses ran WordPress on shared hosting: updates lagged behind, there was nowhere to test changes, and when a site broke nobody knew what had changed. For agencies, every outage hurt their own client relationship. I standardised server setup in half-automated Ruby runbooks with Ansible. Every site gets its own virtual server in an EU data centre plus an identical staging copy. Core, theme and plugin updates run automatically, and Restic/Autorestic take incremental snapshots of database and files. Monitoring covers external uptime (Gatus), scheduled jobs (Healthchecks), metrics (Prometheus), logs (Loki) and dashboards (Grafana). One person operates 20 sites. A new server is ready in 30 minutes instead of 8 hours, daily backups are kept for the whole subscription, and 99% of incidents in 10 years were caught by monitoring before the client noticed.
Exhibition IT for museums and the EU Council Presidency
2016 - 2024
Whyservices OÜ, with a data visualisation studio · Tallinn, Brussels
Industry: museums, public sector (EU2017) · Role: infrastructure: hardware, automation, operations · Team: 4 people at the studio, plus museum staff
Each exhibition runs up to 50 Windows terminals and servers for years, in museums without their own IT staff. Opening dates are fixed, and every terminal that goes dark is visible to visitors. I wrote the hardware specifications for the tenders, weighing minimum requirements against the look the designers wanted. I set up network and hardware and configured the Windows terminals with Ansible playbooks. CI/CD pipelines build the exhibition software and deploy it, along with digital assets, to all terminals. For remote operations I added Raspberry Pi out-of-band units, Wake-on-LAN, VPN access (WireGuard, Tailscale) and monitoring. For EU2017 I coordinated partners in Tallinn and Brussels. Five exhibitions delivered: ERM "Digital Encounters" (2016), EU2017 "Flow" (2017), Fat Margaret Museum (2019), "Heaps of Happiness" at the Estonian Health Museum (2021) and Seaplane Harbour (2024). 90% of incidents are solved remotely, without a site visit.
Head of System Operations
01/2013 - 07/2017
Cash on Go Ltd · Tallinn
Industry: fintech, regulated (PCI DSS, FCA) · Role: head of system operations · Team: 3 system administrators
The company's main application and database ran on hand-administered servers in the data centre and in the office. A hardware fault would have stopped the business, and as a regulated fintech the company had to pass PCI DSS and FCA-related audits. I led the team of system administrators and moved administration to infrastructure as code, first with Chef, later with Ansible. Together with the development team and external consultants, I migrated the PHP/Symfony application and its MySQL database to a redundant, multi-location setup on AWS (EC2, VPC, S3), with internal services on Proxmox. As primary lead for PCI DSS procedures, I introduced standardised work processes and release policies, and was the main contact for vendors and IT service providers. The core application no longer had a single point of failure, and servers could be rebuilt from code. 3 PCI DSS audits passed without major findings. Received the company's 2014 Einstein Award for work across departments.