MA•26
000
DEVOPS & PLATFORM ENGINEER — BUDAPEST

I BUILD & AUTOMATE THE CLOUD PLATFORMS AI RUNS ON.
FR 001 / 120 SCROLL TO ORBIT ↓
THE NUMBERS
0+
Microservices running on EKS & AKS infrastructure I built
0MIN
Environment provisioning, down from 6 hours, via Terraform modules
0+
Production metrics & alerts shipped with Prometheus and Grafana
0
Engineering teams on self-service CI/CD templates I authored
0×
Star of the Quarter awards at Tata Consultancy Services
01
01 — WHAT I DO

The Architect

Multi-cloud, high-availability infrastructure on AWS and Azure. VPCs, EKS, AKS, serverless. Designed to survive failure, audits, and Monday mornings.

02 — WHAT I DO

The Automator

Terraform, Terragrunt, GitOps, blue-green and canary pipelines. Platforms that provision themselves in 45 minutes instead of 6 hours.

03 — WHAT I DO

The Builder

AI platforms and production MCP servers. Machine-learning infrastructure and agents wired into real engineering workflows, not demos.

SELECTED WORK

PLATFORMS
THAT SHIP

↗
W.01

ML Platform

Multi-tenant machine-learning platform on SageMaker Unified Studio. Six-plus data squads onboarded to self-serve tooling.

AWS · SAGEMAKER · TERRAFORM
THE PROBLEM

Data squads queued behind ops for every environment. Notebooks, models and permissions scattered across accounts, nothing reusable.

WHAT I BUILT

A multi-tenant SageMaker Unified Studio platform: Terraform-provisioned per-squad domains, IAM isolation, shared registries, self-serve onboarding.

IMPACT

Six-plus data squads onboarded and shipping without ops in the critical path. One paved road instead of six snowflakes.

SQUAD ASQUAD BSQUAD CSAGEMAKER UNIFIED STUDIOTERRAFORM MODULESS3 DATA LAKE domains · iam isolation · model registry
↗
W.02

Production MCP Servers

MCP servers deployed to production, wiring AI agents into engineering workflows across the organisation.

MCP · AI AGENTS · PYTHON
THE PROBLEM

AI agents could reason but not act: no safe, governed path from an LLM to real engineering systems.

WHAT I BUILT

Production MCP servers exposing engineering tooling through typed tool schemas, scoped credentials and guardrails, deployed like any other service.

IMPACT

Agents joined the daily workflow across teams: querying systems, driving pipelines, adopted org-wide as a pattern.

AI AGENTSCLAUDE · LLMSMCP LAYERCI/CDOBSERVABILITYDATA APIS auth · tool schemas · guardrails
↗
W.03

IaC Module Library

One reusable Terraform and Terragrunt library powering 15+ microservices across AWS and Azure. 95% of config drift, gone.

TERRAFORM · TERRAGRUNT · CI/CD
THE PROBLEM

Every service hand-rolled its own infrastructure: six-hour provisioning, drift everywhere, no two environments alike.

WHAT I BUILT

A versioned Terraform + Terragrunt module library with DRY environment configs and CI validation on every change.

IMPACT

Provisioning down from 6 hours to 45 minutes, 95% of config drift eliminated, 15+ microservices on one paved road across AWS and Azure.

MODULE LIBRARYENV · DEVENV · STAGEENV · PROD15+ MICROSERVICES vpc · eks · rds · iam · cdneks · aks · ecs fargate terragrunt keeps every env identical, ci validates every change
↗
W.04

CDC Data Platform

Org-wide change-data-capture platform on AWS DMS feeding live Power BI reporting for the whole business.

AWS DMS · ETL · POWER BI
THE PROBLEM

Business reporting trailed production by a day, stuck behind overnight batch loads from the core databases.

WHAT I BUILT

An org-wide change-data-capture pipeline on AWS DMS: source databases streamed continuously into an analytics store feeding Power BI.

IMPACT

Reporting went live-on-change for the whole business. The batch window stopped being the bottleneck.

ORACLE · RDSSOURCESAWS DMSCDCS3 STAGEPOWER BI change streams, not batch windows — reporting is live
THE PIPELINE — CAREER AS CODE

EVERY STAGE
GREEN

✓
2018–2022 · BUILD

BSc Computer Science

Al Akhawayn University, Morocco. Abdulla Al Ghurair STEM Scholars award.

✓ PASSED
✓
2022 · TEST

First Prod Deploys

Software Engineer Intern, Ilem Group. Blue-green deploys to Azure via GitLab CI, deployment errors cut 80%.

✓ PASSED
✓
2022–2024 · PACKAGE

MSc Computer Science

University of Debrecen, Hungary. Stipendium Hungaricum scholar. Thesis: smart-home IoT platform with ML automation.

✓ PASSED
▶
2023–NOW · DEPLOY

DevOps & Platform Engineer

TCS Budapest, on Europe Assistance. Multi-cloud production platforms, AI infrastructure, 3× Star of the Quarter.

● RUNNING
?
NEXT · RELEASE

Your Platform Team

DevOps, Platform, SRE, or Data/AI platform roles. The pipeline is green and waiting.

◌ APPROVAL PENDING
CREDENTIALS — DON'T TRUST, VERIFY
AWS
AMAZON WEB SERVICES

AWS Certified Solutions Architect — Associate

VALIDATION Nº aa58ad16a9cf48779fc0869d81235096
ISSUED FEB 2026 · VALID THRU FEB 2029
VERIFY ↗
MICROSOFT × GITHUB

GitHub Copilot Certification

CREDENTIAL ID 32A56DFBA55AFD0D · Nº 346339-9A3845
ISSUED JAN 2026 · VALID THRU JAN 2028
VERIFY ↗
Microsoft Certified Azure Fundamentals badge
MICROSOFT

Azure Fundamentals (AZ-900)

EARNED 2025 · NO EXPIRY
VERIFY ↗
Cisco
CISCO

CCNA — Enterprise Networking, Security & Automation

EARNED 2024 · NETWORKING FUNDAMENTALS
VERIFY ↗
THE TERMINAL — TALK TO THE STACK
visitor@ma26: ~
visitor@ma26:~$ 
AFTER HOURS

STILL
HUMAN

$ crontab -l · 0 18 * * * ./after_hours.sh
Football

On the pitch and in the stands. Ask about tactics, lose an hour.

AI & Tech

New models, new tooling, tested the week they ship.

Film & Photo

Framing, lighting, grading. The cinematic bug is real.

Gaming

Controller in hand. Competitive when it matters.

Guitar

Sometimes. The neighbors survive.

WANT A PLATFORM
THAT SHIPS ITSELF?

Get in touch See the work Download CV ↓