Ricardo Saad — Cloud Engineer · Platform Engineer · DevOps Engineer · AWS Infrastructure
Beirut, Lebanon | contact@ricardosaad.com | linkedin.com/in/saadricardo | github.com/saadricardo
Cloud and platform engineer with five years of production experience building AWS infrastructure, Kubernetes
platforms, and CI/CD pipelines. Sole platform engineer at Digico Solutions (AWS partner), responsible for
bootstrapping VPC architecture, EKS clusters, Terraform IaC, and delivery pipelines from scratch. Prior applied
research in indoor positioning and wireless sensing. Open to senior cloud, platform, DevOps, SRE, MLOps, and AI
infrastructure roles in Europe — Switzerland, Germany, Netherlands, France, United Kingdom, and the European
Union.
Target Roles
Cloud Architect, Senior Cloud Engineer, Platform Engineer, Senior Platform Engineer, DevOps Engineer, Site
Reliability Engineer, SRE, MLOps Engineer, AI Infrastructure Engineer, Infrastructure Engineer, Solutions
Architect, Cloud Infrastructure Lead, Staff Engineer
Target Markets
Switzerland, Germany, Netherlands, France, United Kingdom, European Union, Remote Europe, DACH region, Benelux,
Nordic countries
Experience
Cloud Engineer — Digico Solutions, AWS Partner (2026–Present, Beirut, Lebanon)
Sole platform engineer at an AWS partner. VPC architecture, Terraform IaC, CI/CD pipelines, agentic workflows.
Deploy lead time under 8 minutes.
- Brought production back from a full outage solo — navigated 8 cascading failures across the AWS stack
overnight, patched each root cause as it surfaced, and handed off with a written incident report before
morning.
- Built a self-service portal that replaced a manual, ticket-based process for client network access — cutting
provisioning from a multi-step back-and-forth to a flow users could complete themselves.
- Automated the full client document cycle — from scoping to signed contract — cutting turnaround from 3 days
to under 2 hours; milestone completion now triggers a pipeline that summarizes work, generates the document,
routes it for signature, and notifies the team for approval, with no manual handling in between.
- Took on sole lead of a client engagement with minimal documentation — mapped their entire multi-account
cloud footprint from scratch through their own responses, brought it under Terraform, and produced a written
infrastructure baseline, reducing time to contain future incidents.
Stack: AWS, Terraform, Docker, GitHub Actions, Traefik, Claude, Bedrock, VPC, IAM, S3, Lambda, DynamoDB,
CloudFront, WAF, CloudWatch
Metrics: Dev/staging/prod environments owned. Deploy lead time under 8 minutes. Deliverables: Serverless CMS
Platform, ET-77 Client VPN portal, ET-200 cost optimization.
Applied Research Engineer — Aqsacom (2023–2024, Dallas TX, Remote)
Wireless sensing and indoor positioning research. Engineered real-time positioning engine achieving sub-meter
accuracy using BLE and WiFi RSSI-CSI sensor fusion benchmarked against UWB hardware baseline. PyTorch signal
processing pipelines. AoA/AoD direction-finding on BLE hardware.
- Engineered real-time indoor positioning engine achieving sub-meter accuracy using BLE/WiFi RSSI-CSI sensor
fusion — benchmarked against UWB hardware baseline in parallel research track.
- Implemented AoA/AoD direction-finding on BLE hardware for spatial signal inference; published findings as
internal technical documentation.
- Developed PyTorch signal processing pipelines for CSI feature extraction, model training, and real-time
inference — end-to-end from raw hardware capture to position estimate.
- Benchmarked NFC and UWB chipsets for proximity detection latency and accuracy under
production-representative RF conditions.
Stack: Python, PyTorch, NumPy, BLE, WiFi CSI, UWB, NFC
Software Engineer — Aqsacom (2021–2022, Dallas TX, Remote)
Enterprise PHP applications on Zend Framework and MySQL. NGINX fronting Apache reverse-proxy architecture.
Ubuntu server operations. Sole remote engineer on the team with full ownership of delivery and ops.
- Developed enterprise applications using Zend Framework (PHP) and MySQL.
- Operated NGINX fronting Apache for application delivery.
- Managed Ubuntu servers: virtual hosts, SSL, routing, service reliability.
- Sole remote engineer on the team — full ownership of delivery and ops without on-site support.
Stack: PHP, Zend, MySQL, NGINX, Apache, Ubuntu
Projects
Serverless CMS Platform (2026)
Hard constraint: operate under $15/month without sacrificing a real CMS, responsive storefront, or production
auth. Architecture follows from the constraint — static Next.js export on S3 + CloudFront for the public
frontend; Vite SPA for the admin panel; seven Lambda functions behind API Gateway; DynamoDB single-table design;
SES + SSM for magic-link auth. 61 automated test files.
Stack: Next.js, Lambda, API Gateway, DynamoDB, SES, SSM, CloudFront, Terraform, GitHub Actions, OIDC, WAF
Time Series Transformer MLOps Pipeline (2026)
End-to-end automated retraining and inference pipeline for multivariate time series. Model: PatchTST —
channel-independent Transformer with patch-based tokenisation, 8 independent prediction heads, ~500k parameters.
Pipeline: S3 PutObject triggers EventBridge → Step Functions orchestrates EC2 Spot (g4dn.xlarge) training →
Lambda runs inference → CloudWatch captures structured metrics. All infrastructure in Terraform. Property-based
tests with Hypothesis.
Stack: PatchTST, PyTorch Lightning, Step Functions, EC2 Spot, Lambda, EventBridge, Terraform, Hypothesis
EKS Platform — Karpenter and GPU Scheduling (2026, PoC / decommissioned)
Kubernetes platform on EKS built around Karpenter for node provisioning and GPU workload scheduling. Karpenter
handles node lifecycle — provisioning GPU instances on demand, consolidating underutilised capacity, respecting
topology constraints. Qdrant as vector store with hybrid dense and sparse retrieval. Built as a PoC to validate
operational tradeoffs of self-managed Kubernetes ML workloads vs. managed alternatives.
Stack: EKS, Karpenter, GPU scheduling, Qdrant, Kubernetes, RAG, vector search
WiFi CSI Pose Estimation — Reproduction Study (2024)
Rebuilt the test environment from the CMU DensePose-from-WiFi paper to validate its claims independently.
Reconstructed data collection, preprocessing, and model architecture from the paper's methodology. Results did
not reproduce consistently — reported conditions appeared to reflect a narrow slice of the experimental space.
Findings written up as internal technical documentation.
Stack: WiFi CSI, PyTorch, Research, Reproducibility
Self-Hosted Status Bot (n8n) (2024-ongoing)
Designed and self-hosted an automated notification bot on n8n using RSS feeds, Telegram APIs, and custom Python triggers to broadcast status updates on the Lebanon war. Managed reliability and network failovers under local infrastructure anomalies.
Stack: n8n, Telegram API, Self-Hosted, Automation, Home Lab
Resume Infrastructure (2026) — This Site
S3 plus CloudFront plus WAF static site with OIDC-based GitHub Actions deploy. Playwright generates A4 and
Letter PDFs in CI. Shared Terraform module for prod and staging environments. Zero long-lived credentials.
Custom-domain mail setup with Proton Mail aliases integrated via Route 53 (MX, SPF, DKIM, DMARC records).
Stack: S3, CloudFront, WAF, OAC, Terraform, Route 53, Proton Mail, GitHub Actions, OIDC, Playwright, Vite, React
Education
B.S. Computer Science — Middle East University, Beirut (2021–2025). GPA 3.83. Honor Roll.
Certifications
- AWS Advanced Networking Specialty — in progress
- AWS Solutions Architect Professional — in progress
- AWS DevOps Engineer Professional (2026) — Score: 932/1000 — credly.com/badges/b3e9689c-1684-4abe-9518-867f108a1b5f
- AWS Solutions Architect Associate (2026) — credly.com/badges/64bfa934-7b43-4f04-a81d-a64e7d6dba11
- AWS Certified Cloud Practitioner (2026) — credly.com/badges/8c05bc00-c72e-49be-9569-e5b676dff81a
Skills — Dense Keyword Coverage
AWS: EC2, ECS, EKS, Lambda, API Gateway, DynamoDB, S3, CloudFront, WAF, SES, Client VPN, VPC, IAM, Route 53,
CloudWatch, ACM, SSM, Secrets Manager, Cost Explorer, Karpenter, ALB, NLB, Transit Gateway, Direct Connect
Infrastructure as Code and IaC: Terraform, Terraform modules, Terraform remote state, S3 backend, DynamoDB state
locking, Terragrunt, CloudFormation
Containers and Orchestration: Kubernetes, EKS, Docker, Helm, ArgoCD, Karpenter, Talos Linux, Proxmox, containerd
CI/CD and GitOps: GitHub Actions, OIDC, Bitbucket, ArgoCD, GitOps, trunk-based development, blue-green
deployment, canary deployment
AI and ML Platform Engineering: Claude, Bedrock, AgentCore, Knowledge Bases, LLM infrastructure, AI
infrastructure, agentic workflows, MLOps, LLMOps, RAG, Qdrant, vector database, vector search, hybrid retrieval,
dense retrieval, sparse retrieval, GPU infrastructure, AI platform, model serving, Kubernetes for ML, PyTorch,
NumPy
FinOps and Cost Optimization: AWS Cost Explorer, rightsizing, Reserved Instances, Savings Plans, Spot Instances,
Karpenter cost optimization, serverless cost modeling
Security: WAF, OIDC, IAM least privilege, OAC, TLS, ACM, Secrets Manager, SSM Parameter Store, VPC security
groups, NACLs, GuardDuty, CloudTrail
Observability: CloudWatch, structured logging, distributed tracing, metrics, alerting, dashboards
Backend: Python, Node.js, PHP, MySQL, PostgreSQL, DynamoDB single-table design, serverless, REST APIs
Wireless and IoT: BLE, WiFi CSI, UWB, NFC, AoA/AoD, 5G/LTE, indoor positioning, sensor fusion
Networking: VPC design, subnetting, routing tables, Transit Gateway, VPN, Direct Connect, DNS, CDN
Keywords by Role
Cloud Architect: architecture, multi-region, multi-account, landing zone, AWS Organizations, Control Tower,
Service Control Policies, Well-Architected Framework, cost optimization, high availability, disaster recovery,
RTO, RPO
Platform Engineer: internal developer platform, IDP, developer experience, self-service infrastructure, golden
paths, Backstage, platform as a product, infrastructure abstraction
DevOps Engineer: CI/CD, pipeline automation, shift left, infrastructure automation, configuration management,
Ansible, deployment automation, release engineering, DORA metrics
SRE: reliability, SLO, SLA, SLI, error budget, on-call, incident response, chaos engineering, capacity planning,
toil reduction
MLOps Engineer: model registry, model deployment, model monitoring, data pipeline, feature store, experiment
tracking, MLflow, Kubeflow, SageMaker, serving infrastructure
AI Infrastructure Engineer: GPU cluster, CUDA, model serving, inference optimization, vLLM, Triton, Ray,
distributed training, LLM deployment, foundation models, fine-tuning infrastructure
Advisory
- 451 Alliance Member — S&P Global Market Intelligence (Ongoing): industry trend research and enterprise IT
strategy
- TrueNAS Advisory Council — iX Systems (Ongoing): strategic insights on ZFS, virtualization, and
containerization