Florin

Ungur

github.com/florinungur  |  florinungur.com  |  florin@florinungur.com

Competent and results-oriented SRE with deep experience in managing critical infrastructure across colocation providers and all major cloud platforms. Proven ability to implement automation and monitoring solutions on every leading container orchestrator. Passionate about supporting vital sectors such as water, energy, and commerce.

I pledge to serve humanity by making sure that software does what it should. I pledge to keep information available to anyone who needs it and unavailable for any unauthorised party who wants it. I pledge to make our systems more performant, more secure, and more beautiful. To achieve these pledges I promise to become a better man by contributing to society and human wellbeing in an honest and trustworthy way while avoiding doing harm.

Experience

Apple

Site Reliability Engineer, ML | July 2024 – Present • <1 year | London, UK

Let's see...

Blockchain.com

Site Reliability Engineer | Mar 2022 – July 2024 • >2 years | London, UK

Weathered multiple crypto winters and summers, contributing to the company's $6B evaluation through dozens of initiatives. Main projects:

  • Eliminated Exchange time drift: Collaborated with our network engineer to implement NTP on leaf switches and automated Chrony NTP configuration using Tinkerbell, ensuring a more robust and resilient trading experience.
  • Led critical migration: Successfully migrated the most crucial database from MySQL 5.7 to MySQL 8.
  • Improved service restarts: Extended the Golang deployment tool to enable restarts of individual service components, minimising trading downtime for the Exchange and reducing the impact of deployments.
  • Empowered teams with ownership: Automated the Grafana-based alerting system so that each team receives alerts only for the services that they own, improving their focus and response times.
  • Strengthened access management: Collaborated with a teammate to successfully integrate GitHub and Kafka into the Python-based permission system, strengthening security controls.
  • Revolutionised support: Established the SRE helpdesk, saving ~200 dev hours per month.

Tech stack: Python Golang Bash / GCP AWS On-prem Debian / HashiStack Kubernetes Traefik Confluent / Spinnaker Concourse GitHub Actions / Terraform Ansible SaltStack / Prometheus Grafana Datadog / PostgreSQL MySQL Redis

GreenFlux

DevOps Engineer | Jun 2021 – Mar 2022 • 10 months | Amsterdam, NL

Together with my tech lead we optimised the entire infrastructure needed to serve thousands of customers across four continents. In particular:

  • Contributed to the EV charging point infrastructure: Played a pivotal role in creating the FTP solution used by tens of thousands of charge points to upgrade their firmware.
  • Created a new monitoring platform: Refactored and revamped the monitoring system, enhancing the developer experience and significantly reducing costs.
  • Mitigated high-risk vulnerabilities: Successfully addressed the ChaosDB, OMIGOD, and CredManifest CVEs.

Tech stack: C# PowerShell Bicep / Azure / Elasticsearch Kibana

Thales Nederland

Infrastructure (Security) Engineer | Feb 2020 – Jun 2021 • <2 years | Hengelo, NL

I am proud to say that I have played a part in transitioning TACTICOS, the operating system of battleships operated by tens of military forces around the world, from an Ada-based orchestrator to Kubernetes. Key contributions:

  • Proved that Kubernetes is military-grade: Created Java Cucumber tests for Kubernetes that proved the orchestrator's robustness in Thales' environment: github.com/florinungur/RCNIT + github.com/florinungur/bashspray.
  • Facilitated legacy migration: Transitioned legacy software from a RHEL 7 to a RHEL 8 infrastructure.
  • Enhanced automation and decoupling: Created Ansible roles to automate and decouple configuration that was previously done by critical software.

Tech stack: under NDA

More working experience can be shared upon request.

Education

Saxion University of Applied Sciences

Bachelor's Degree in IT Service Management | Nov 2016 – Jul 2020

  • › Graduated cum laude with an average of 8.4/10; top of the 2020 class.
  • › Note-worthy modules: Architecting the Cloud, Cloud Engineering, Servers & Services, Networking.
  • › Recommended for my work by Tonny Lievers.

University of Twente

Minor in Psychology | Feb 2019 – July 2019

  • › Graduated with an average of 8/10; top 5% of the class.
  • › Focused on linguistics, cognition, IQ, and the psyche for the bringing together of IT security and psychology, as inspired by the second edition of Ross Anderson's book, Security Engineering.

Certificates

AWS Academy Cloud Foundations

September 26, 2019

AWS Academy Cloud Architecting

October 3, 2019

Notable projects

flaregun

github.com/florinungur/flaregun | Last updated: January 14, 2024

A website for increasing security awareness.

hfi

github.com/florinungur/hfi | Last updated: August 16, 2023

hfi (short for "how fast is") is a wrapper over the time(1) command that shows the time, CPU, and memory use of a Bash script in a pretty format.

Docker permission errors are not ok

florinungur.com/essays/2023/08/12/docker-permission-errors-are-not-ok | August 12, 2023

An investigation of the way in which Docker handles permissions inside and outside a container.

/usr/local/bin/polkadot: line 1: Not: command not found

florinungur.com/essays/2022/04/18/not-command-not-found | April 18, 2022

A deep dive into how the Polkadot cryptocurrency handles new software releases.

Azure autoscaling best practices: an addition

florinungur.com/essays/2021/07/25/azure-autoscaling-best-practices-an-addition | July 25, 2021

The story of how I found a bug in Azure's autoscaling mechanism.

IlRlY2hub2xvZ3kncyBqdXN0IG5hdHVyZSB3ZSB0YXVnaHQgdG8gZG8gY29vbCB0cmlja3MuIiAtIGV4dXJiMWEK