Vedat Y.

I build infrastructure and automation systems that run reliably in production — not prototypes, not proofs of concept. With 15+ years across cloud, on-premises, IoT, and network environments, I solve operational problems with clean engineering and documented, maintainable code. I work across three focused service lines: ☁️ Cloud Infrastructure & IoT Systems I design and operate AWS environments that stay up under real-world conditions. My most recent engagement was a 5-year contract with a Norwegian IoT company — 100+ connected devices across 50+ distributed sites, maintained at 99.9% uptime with zero security incidents. What I deliver: AWS architecture (EC2, VPC, S3, IAM, IoT Core) — designed for reliability and cost efficiency Zero-downtime migrations from on-premises to cloud IoT fleet management (Balena, AWS IoT Core, MQTT, embedded Linux, Raspberry Pi) Full observability stack: Prometheus, Grafana, Zabbix — with automated alerting and self-healing Results from production: 40% MTTR improvement, 60% reduction in manual operational tasks, 75% reduction in IoT device connectivity issues. 🔧 Network Automation & Infrastructure-as-Code I automate network infrastructure the right way — with a source of truth, not ad-hoc scripts. My approach: NetBox as the authoritative data source, Terraform and Ansible for provisioning and configuration management, Python tooling for operational workflows. Hands-on experience with: MikroTik RouterOS and CHR (including physical-to-cloud migration) VPN architecture, VLAN design, firewall rules, QoS, per-user traffic management RouterOS scripting for automated network operations Ansible playbooks across 50+ distributed Linux servers and network devices ⚙️ Python Automation & Scheduled Services I build Python services that replace manual workflows — properly engineered, containerised, and deployed in production. Recent example: a scheduled SEO monitoring service built entirely in Python (APScheduler, FastAPI, httpx, Docker Compose), deployed on a VPS. It runs weekly scans across client websites using a custom 13-category scoring engine (PageSpeed Insights API, CrUX API, custom crawler — no third-party SEO data costs), fires Slack and email alerts when scores drop below defined thresholds, and auto-remediates common issues directly at the CMS data layer (WordPress/RankMath via WP-CLI and REST API). Claude API is used narrowly for plain-English alert summaries and monthly digests — the scoring and remediation logic is pure Python. This is the kind of automation I build: scheduled, reliable, observable, and genuinely useful.