Founder · Platform & Observability Engineer

Mehrshad Lotfi.

I help teams run Kubernetes platforms that are observable, reliable and cost-efficient.

Grafana Champion and Kubestronaut with a research background in cloud networking. I design and operate the Grafana LGTM stack, Prometheus and OpenTelemetry on Kubernetes, and I turn monitoring into code your team can own.

Frankfurt am Main · remote or on-site within 250 kmAvailable for new projects

Grafana Champion 2024KubestronautUpwork Top Rated Plus

Book a call

Founder & Platform EngineerOptiop GmbH

$150k+

yearly cloud savings delivered for one client

~20%

lower cloud spend after moving off CloudWatch

15+

cloud-native certifications

100%

Job Success Score on Upwork

Companies I've worked with.

From exchanges and banks to research and fast-growing SaaS teams

Deutsche Börse

Software / DevOps engineer

dwpbank

Monitoring & alerting

MPI-SWS

Max Planck Institute for Software Systems

Healthcare

Cloud communications

Technology partner

ICT solutions

Selected projects.

What I built, and what it changed for the teams I worked with

Healthcare SaaS · via Optiop2025

$150k+

saved per year

From CloudWatch to the Grafana LGTM stack

Replaced CloudWatch-based logging, metrics and tracing with a self-hosted stack on Kubernetes: Loki for logs, Tempo for traces and Cortex for long-term metrics, all visualised in Grafana and instrumented with OpenTelemetry.

Cut the client's cloud spend by around 20%, more than $150k per year
One place to correlate logs, metrics and traces during incidents
Rolled out with Helm and ArgoCD so every change is reviewed in Git

GrafanaLokiTempoCortexOpenTelemetryAWSArgoCD

E-commerce & healthcare · via Optiop2025 – today

Dedicated monitoring platform with SSO

Built dedicated monitoring clusters that run independently of the workloads they observe, deployed with Helm and ArgoCD. Dashboards and alert rules live in Git as code, and access is handled through single sign-on.

Alerts and dashboards as code, versioned and reviewed like application code
SSO with Azure Entra ID and Keycloak, mapped to team roles
Monitoring stays available even when production clusters are degraded

KubernetesHelmArgoCDPrometheusGrafanaKeycloakEntra ID

Platform teams · via Optiop2025

Cost and delivery visibility with OpenCost & DORA

Introduced OpenCost to break Kubernetes spend down by team and workload, and DORA metrics to measure how fast and how safely teams ship.

Per-namespace and per-team cost dashboards in Grafana
Deployment frequency, lead time, change failure rate and MTTR tracked automatically
Shared numbers for engineering and management discussions

OpenCostPrometheusGrafanaKubernetesDORA

Deutsche WertpapierService Bank (dwpbank)

Monitoring and alerting for banking microservices

Worked as a DevOps engineer improving monitoring and alerting across microservices, databases and Kafka, with a focus on actionable alerts and faster root-cause analysis.

Consistent dashboards for services, databases and Kafka
Alert rules tuned to reduce noise and catch real incidents earlier

PrometheusGrafanaKafkaMicroservicesDatabases

Deutsche Börse02/2023 – 07/2023

High-performance client for Cloud Stream market data

Built a high-performance client for Deutsche Börse's Cloud Stream market data service and supported the migration of workloads from AWS to Google Cloud.

Low-latency consumption of real-time market data
Hands-on support for the AWS to Google Cloud migration

Google CloudAWSStreamingPerformance engineering

Max Planck Institute for Software Systems10/2019 – 01/2023

CPU-efficient network stacks for cloud virtualization

Research on making cloud network virtualization cheaper in CPU cycles, working close to the hardware with virtual switches, NICs and offloading.

Research on CPU-efficient networking for multi-tenant clouds
Deep understanding of how networking behaves under load, which now shapes how I monitor it

Open vSwitchNICsPCIeFPGAC/C++

Freelance highlights

A selection of completed projects on Upwork

Real-time alerts for a crisis management application

Real-time alerting and notifications built on Grafana dashboards and Loki log queries.

GrafanaLokiAlerting

Grafana Cloud monitoring for network appliances

Monitoring of network appliances with Grafana Cloud, from data collection to dashboards.

Grafana CloudNetworking

Velocity dashboard

Dashboard build project in Grafana, delivered end to end and rated five stars by the client.

GrafanaDashboards

AWS infrastructure & application monitoring

End-to-end monitoring for AWS infrastructure and the applications running on it.

AWSGrafanaPrometheus

Experience.

2025 – today
Founder & Platform Engineer
Optiop GmbH
Observability, platform engineering and Kubernetes consulting for teams in healthcare and e-commerce.
08/2023 – today
Freelance DevOps & Site Reliability Engineer
Self-employed
Observability and Kubernetes projects for clients in Germany and worldwide, including via Upwork.
02/2023 – 07/2023
DevOps Engineer
Deutsche Börse
Cloud Stream market data client and AWS to Google Cloud migration.
10/2019 – 01/2023
Research Associate
Max Planck Institute for Software Systems
CPU-efficient network stacks for cloud network virtualization.

Education

M.Sc. Computer Science
Saarland University & Max Planck Institute for Software Systems
B.Sc.
Sharif University of Technology

Languages

English (business fluent)German (business fluent)Persian (native)

Certifications.

Verified credentials across Kubernetes, observability and platform engineering

Kubestronaut

The Linux Foundation · 2025