YM Yahya Mohand
CV Let’s talk

Available for new roles · Open to relocation to Doha, Qatar

I keep payment platforms online — even when a whole region goes down.

I’m Yahya Mohand Isam, a senior cloudops engineer and AWS Certified Solutions Architect. Since 2021 I’ve designed, run and secured production systems on AWS for payment and e-commerce platforms with strict sub-5-minute SLAs.

See the migration story Download CV
3 AWS credentials
data lost moving a live payments platform across regions
0 data lost moving a live payments platform across regions
final production cutover window
~85 min final production cutover window
less downtime from incident fixes
~50% less downtime from incident fixes
faster deployments with reusable CI/CD
~30% faster deployments with reusable CI/CD

Credentials

Certified by Amazon Web Services

Architecture, machine learning and teaching, each verified on Credly. Hover to tilt a badge, and tap it to flip it over for details.

  • AWS Certified Solutions Architect – Associate badge
    Associate

    AWS Certified Solutions Architect – Associate

    Issued June 2022 · Amazon Web Services

  • AWS Certified Machine Learning – Specialty badge
    Specialty

    AWS Certified Machine Learning – Specialty

    Issued July 2023 · Amazon Web Services

  • AWS re/Start Accredited Instructor badge
    Accredited

    AWS re/Start Accredited Instructor

    Issued June 2022 · Amazon Web Services

View all badges on Credly

Case study

Zero-data-loss region migration, under fire

AWS me-south-1 (Bahrain) → eu-central-1 (Frankfurt)

An availability-zone failure destabilised the AWS Bahrain region that hosted a live payment & voucher platform. With a hard 5-minute peak-hours SLA and roughly 72 hours on the clock, I led the emergency move of the whole platform to Frankfurt — driven by a precise 8-phase runbook.

Cutover, simulated

eu-central-1 Frankfurt SECONDARY · ~50 ms lag PRIMARY · WRITER me-south-1 Bahrain PRIMARY · AZ DEGRADED RETIRED
  1. waf · traffic gate closed at the edge
  2. aurora · planned global failover started
  3. aurora · eu-central-1 promoted to primary (~3 min)
  4. origin · internal DNS record → eu-central-1
  5. waf · traffic gate open · data lost: 0
deadline
~72h
final cutover window
~85 min
database promotion
~3 min
replication lag
~50 ms
data lost
0
  1. 01

    Rebuild the region as code

    Recreated the target region with Terraform — VPC, ECS Fargate, ALB, ElastiCache, SQS, ECR and SSM Parameter Store — mirroring production while correcting regional ARNs and resource names.

    TerraformECS FargateVPC
  2. 02

    Replicate at the storage layer

    Built the cutover around Amazon Aurora Global Database: storage-layer replication at ~50 ms lag instead of restoring a stale snapshot.

    Aurora Global DBPostgreSQL
  3. 03

    Run parallel workstreams

    S3 Cross-Region Replication and payment, voucher and third-party provider IP whitelisting ran in parallel, each tracked against a go/no-go checklist.

    S3 CRRRunbooks
  4. 04

    Gate traffic deterministically

    Used a CloudFront WAF WebACL and a single internal-DNS origin record as the traffic gate, so flipping production — or rolling back — was instant and unambiguous, with no waiting on DNS TTLs.

    CloudFrontWAFRoute 53
  5. 05

    Planned failover, zero data loss

    A managed planned failover promoted the Frankfurt cluster to primary in ~3 minutes. Reversible, with no data lost.

    FailoverDR
  6. 06

    Harden after cutover

    Within 24 hours of cutover, removed every hard-coded region reference from the GitHub Actions and CircleCI pipelines.

    GitHub ActionsCircleCI

Experience

Five years on call for production

From consulting on cloud architecture to running payment infrastructure, and teaching the next generation of cloud engineers along the way.

  1. Cloud Engineer / DevOps

    DigitalZone · Iraq · Remote

    Mar 2024 — Present

    • Led the zero-data-loss emergency migration of a live payments platform from Bahrain to Frankfurt.
    • Architected cloud-native infrastructure on ECS, RDS Aurora, Lambda and S3 for performance and reliability.
    • Resolved critical production incidents (database spikes, ECS resource exhaustion), cutting downtime by ~50%.
    • Cut deployment time by ~30% with reusable CircleCI workflows integrated with Terraform.
    • Moved secrets to SSM Parameter Store with KMS encryption, with no hard-coded credentials anywhere.
    • Set up proactive monitoring and alerting with CloudWatch and AWS Chatbot, so incidents get a faster response.

  2. Technical Lead — AWS re/Start

    Mosul Space · Iraq

    Mar 2022 — Present

    • Led the AWS re/Start workforce program: 70% of the first cohort passed AWS Certified Cloud Practitioner.
    • Designed coursework and hands-on labs from real-world cloud scenarios, and mentored career-changers into cloud roles.
    • Delivered professional-skills training and partnered with organisations to broaden inclusion in tech.

  3. Machine Learning Engineer

    Falaq · Mosul, Iraq

    Mar 2023 — Jun 2023

    • Built an RNN (Keras) sales-forecasting model for a platform connecting pharmacies with drug stores.
    • Raised predictive accuracy from 72% to 95% and cut inference latency by ~25%.

  4. AWS Solutions Architect

    Mosul Space · Iraq

    Sep 2021 — Feb 2023

    • Consulted on cloud solution design across compute, virtualisation and enterprise storage.
    • Built disaster-recovery and cloud-backup offerings: cost-effective, compliant off-site protection.
    • Migrated customers to modern virtual server and storage platforms, lowering total cost of ownership.

Selected work

Platforms I’ve built

GitHub

  • Platform engineering

    Ephemeral environments per pull request

    Replaced a traditional staging environment with trunk-based development. Every PR gets an isolated DB schema, ECS service, ECR repo and CloudFront/ALB, with per-PR Terraform state in S3, created and destroyed automatically by GitHub Actions and announced in Slack.

    TerraformGitHub ActionsECSSSMAge

  • Disaster recovery

    Cross-cloud database backups

    Lambda and S3 pipeline that replicates database backups to a second cloud provider, so a failure at one provider never takes the only copy with it.

    LambdaS3Python

  • Data engineering

    Self-hosted data platform

    RudderStack, Airbyte, Airflow and ClickHouse on S3, provisioned entirely with Terraform to cut cost and keep customer data private.

    AirflowClickHouseAirbyteRudderStack

  • MLOps

    End-to-end ML pipeline

    SageMaker training with MLflow tracking, model artifacts on S3, and batch inference orchestrated by Airflow.

    SageMakerMLflowAirflowS3

Toolbox

The stack I run in production

Cloud & Compute

  • AWS
  • ECS Fargate
  • EKS
  • Lambda
  • EC2
  • Aurora PostgreSQL
  • Aurora Global DB
  • ElastiCache / Redis
  • SQS
  • S3
  • CloudFront
  • Route 53
  • WAF
  • ECR
  • API Gateway

IaC & Automation

  • Terraform
  • GitHub Actions
  • CircleCI
  • Docker
  • Trunk-based development
  • Ephemeral environments

Reliability & Security

  • Incident response
  • On-call
  • Cross-region DR
  • SSM Parameter Store
  • KMS
  • IAM / IRSA
  • Cognito
  • Age encryption
  • VPC networking

Observability

  • CloudWatch
  • AWS Chatbot
  • Grafana
  • Prometheus
  • Alerting
  • Go/no-go runbooks

Data & MLOps

  • Airflow
  • ClickHouse
  • Airbyte
  • RudderStack
  • SageMaker
  • MLflow

Languages

  • Python
  • Bash
  • SQL
  • YAML
  • Terraform HCL

Education

B.Tech. in Computer Engineering

Northern Technical College · Mosul, Iraq

2017 — 2021 · GPA 3.7

Awards

  • Agriculture Hackathon

    Qaf Lab · 2023

  • Best Graduation Project

    Northern Technical College · 2021

Contact

Need someone to keep your platform up?

I’m open to Cloud, DevOps and SRE roles, remote or on-site. Open to relocation to Doha, Qatar.

yahyamohandessam@gmail.com

Mosul, Iraq