Freelance practice opening soon
Platform Engineer and SRE for hire

JérémyDelgado

I make infrastructure boring: reliable, cheap and quiet at 3 am. Proven on a platform serving 8M requests per second.

8M
requests / second, worldwide
4,500+
servers in production
30+
Kubernetes clusters
50 TB
PostgreSQL, 24/7
01

What I can do for you.

Infrastructure & SRE

Stop firefighting. I set up SLOs, error budgets and zero-downtime deploys on bare metal or cloud, so incidents become rare and boring.

AnsibleTerraformHAProxySLO / error budgets

Observability

Know before your users do. Prometheus, Thanos and Grafana with alerts that mean something, and an on-call that does not burn your team out.

Kubernetes & GitOps

Every change reviewed, reproducible and reversible. GKE, EKS or on-prem clusters, driven by ArgoCD.

FinOps

Cut the cloud bill without cutting reliability. Rightsizing, spot and committed use, measured before and after.

RightsizingSpot / PreemptibleReserved / CUDsBudget alerts
02

Recent work.

2022
Oct 2022 / present

Equativ

Senior Platform Engineer

Global adtech platform, 8M requests per second across EU, NA and APAC. The challenge: keep a hybrid bare-metal and cloud fleet reliable 24/7 while traffic grows. What I delivered:

8Mreq / s
4,500+servers
~30K8s clusters, GitOps
50 TBPostgreSQL
  • Fleet management: Ansible + Git on OVH, LeaseWeb, AWS
  • Rancher + ArgoCD for on-prem K8s GitOps
  • Prometheus, Thanos, AlertManager, Oncall
  • Terraform on GCP (GKE, CloudSQL, GCS) and AWS (EKS, RDS, S3)
  • Confluent Kafka: KSQL, Schema Registry
  • Incident response, on-call rotation, postmortems
2020
Sep 2020 / Jul 2022

Waays

Web Architect

Multi-cloud hosting provider for digital agencies and e-commerce. The challenge: host dozens of client sites across OVH, Scaleway, Vultr and DigitalOcean without downtime. What I delivered:

700servers
4cloud providers
  • Multi-provider Kubernetes design and operations
  • MariaDB performance tuning for high-traffic workloads
  • GitOps CI/CD for dozens of client environments
  • Zero-downtime live migrations across providers
  • Infra sizing, cost optimization, backups
  • HAProxy load balancing and SSL termination
2017
Education

CPE Lyon, 2020

Engineer, IT & Communication Networks

DUT Réseau & Télécoms, IUT Annecy le Vieux, 2017.

03

Tools I trust.

Cloud native first, GKE at the core. Scroll and they fall into place.

GKE & Google Cloud

  • GKE
  • Workload Identity
  • Cloud SQL
  • Cloud Storage
  • Cloud DNS
  • Cloud Monitoring
  • Terraform / Terragrunt

Cloud Native

  • Kubernetes
  • ArgoCD / GitOps
  • Helm
  • ExternalDNS
  • Rancher / kubeswitch
  • EKS (AWS)

Observability

  • Prometheus / Thanos
  • Grafana / Loki
  • Grafana Tempo
  • OpenTelemetry
  • AlertManager / Oncall
  • SLO / SLI / Budgets

FinOps

  • Rightsizing requests
  • CUDs / Reserved
  • Spot VMs on GKE
  • Budget alerts & labels
  • Cost optimization
04

What I build on my own time.

Small sharp tools, each born from a real annoyance at work.

See more on GitHub
05

Off-call, still busy.

Cooking

A personal cookbook of what I actually cook at home. Reliability is not just for servers.

Open the recipes

50

Claude Code skills

If I do it twice, it becomes a skill.

incident-responsek8s-rightsizingmr-openticklerpost-merge-verifygitlab-job-triageinvestigate-alertgrafana-dashboardstf-module-releaseoncall-triageretrowiki-fact-checkand way more

Sunsets & flowers

Wherever I travel, the camera points at two things: the sky at golden hour and whatever is blooming.

Red poppy in a garden
Chiang Mai, 2022
Sun setting over a lake, its reflection on the water
Jordan, 2023
Red bougainvillea under a blue sky, a butterfly on it
Mexico, 2025
Sun setting behind a mountain peak above a sea of clouds
La Réunion, 2024