Dmitry Zhuravlev
Cloud DevOps, Platform Engineering & AI Infrastructure Portfolio
Production-grade cloud platforms, AI infrastructure, reliability-focused systems, and secure delivery solutions built around operational challenges.
GitHub Developer Program Member
Recognized as a developer building integrations, automation, and tooling on the GitHub platform.
Focus Areas: AI Infrastructure • Distributed Agent Systems • DevSecOps • Platform Engineering • SRE • Cloud Architecture • Kubernetes • CI/CD • FinOps • Infrastructure as Code • Observability • Automation
Featured Platform Projects
1. AI Infrastructure & Operations Platform
GCP-first AI-native operations control plane for governed DevOps, SRE, Platform Engineering, FinOps, security, and cloud operations workflows.
Core Components:
- Python control plane — durable task, attempt, review, and execution workflow management
- PostgreSQL — persistent operational state independent of agent sessions
- Executor adapters — replaceable agent and automation backends behind capability boundaries
- Capability verification — validates executor capabilities before active attempts
- Human review — explicit approval gates for governed operational actions
- GCP Stateful Agent Runtime — persistent private runtime foundation for AI agents
- Secret Manager & IAP — protected secrets and operator access
- GitHub — repository workflows, validation evidence, and auditable engineering changes
What this project delivers:
- Durable orchestration independent of individual AI agent sessions
- Governed execution with explicit human approval boundaries
- Replaceable executors instead of dependency on a single agent framework
- Capability-aware dispatch and controlled retry workflows
- Foundation for coordinated AI-assisted operations across SRE, CI/CD, FinOps, security, and cloud platforms
2. Continuous Integration Build Platform
GCP-based CI platform for on-demand ephemeral self-hosted GitHub Actions runners with a Cloud Run controller and Terraform-managed GCE infrastructure.
Core Components:
- GitHub Actions — runner request and CI workflow orchestration
- GitHub REST API — short-lived runner registration tokens
- Cloud Run — controller for authenticated provisioning requests
- Terraform — GCE runner infrastructure provisioning and management
- Compute Engine — hosts the self-hosted runner
- Docker / Docker Hub — container image build and publishing
What this project delivers:
- On-demand self-hosted runner provisioning
- Ephemeral GitHub runner registration
- Isolated CI execution on GCE
- Automated VM shutdown after workload completion
- Historical end-to-end workflow validation
3. SRE & Reliability Engineering Platform
GitOps-based SRE platform implementing SLO- and error budget–driven release governance with validated canary promotion, automated abort, and recovery workflows on GKE.
Core Components:
- Kubernetes (GKE) — multi-environment runtime for SLO-governed application delivery
- Argo CD — GitOps control plane for declarative application deployment
- Argo Rollouts — canary delivery with SLO-based analysis gates at 10% and 50%
- Prometheus — multi-window SLO, error-ratio, and burn-rate evaluation
- Grafana — visualization of SLOs, error budgets, rollout health, and decision signals
- k6 — deterministic baseline and failure traffic for validating rollout decisions
- Helm — reusable application and platform configuration
- Terraform — modular provisioning of GKE and supporting infrastructure
What this project delivers:
- SLO-gated progressive delivery
- Automated canary promotion or abort based on service-level signals
- Multi-window burn-rate and error-budget evaluation
- Deterministic failure injection and rollout validation
- Operational recovery from failed releases
4. Secure Delivery Platform
Secure cloud-native delivery platform focused on trusted builds, policy enforcement, and Kubernetes-native deployment controls.
Core Components:
- Cloud Build — reproducible and isolated build pipeline for trusted artifact creation
- Artifact Registry — immutable artifact storage with controlled access
- Binary Authorization — policy-based deployment validation ensuring only trusted images are deployed
- Cloud Deploy — progressive delivery with controlled rollout strategies
- Kubernetes (GKE) — controlled runtime environment with enforced deployment policies
- Secret Manager — centralized secret storage with secure injection into workloads
- Terraform — standardized infrastructure provisioning using reusable and customized Google Cloud modules
- Cloud Logging & Monitoring — audit trail, deployment visibility, and operational signals
What this project delivers:
- Controlled software delivery pipeline with integrated policy validation
- Deployment verification and enforcement mechanisms
- Trusted artifact lifecycle from build to runtime
- Security integrated into the delivery workflow
5. Enterprise Cloud Migration
Migration and modernization project focused on architecture, delivery automation, and cloud operating models.
Core Components:
Compute & Runtime
- Compute Engine + Managed Instance Groups — scalable and resilient application hosting layer
- Cloud Run — containerized services for flexible workload execution
Networking
- Custom VPC with private subnets — isolated network environment
- No public IPs on instances — reduced attack surface
- Cloud NAT — controlled outbound connectivity
- HTTPS Load Balancer — secure entry point with TLS termination
Identity & Access
- Dedicated service accounts — workload-level identity isolation
- Granular IAM roles — least-privilege access to Firestore and Cloud Storage
Data & Storage
- Firestore (Datastore mode) — managed NoSQL database
- Cloud Storage — object storage for application data
Infrastructure & Delivery
- Terraform — infrastructure as code with reproducible environments
- GitHub Actions — automated build and deployment workflows
What this project delivers:
- Migration from traditional infrastructure to cloud-native architecture
- Private networking model with controlled access patterns
- Integration of infrastructure, identity, and application delivery
- Multi-layer system design across compute, networking, and CI/CD
6. FinOps & Cloud Cost Assessment
Advanced cloud assessment platform for identifying waste, evaluating optimization opportunities, and producing structured FinOps findings across Google Cloud environments.
Core Components:
- Python — modular assessment engine separating resource discovery, normalization, evaluation, and reporting
- Google Cloud APIs — structured discovery of projects, resources, utilization signals, and billing-relevant metadata
- Cloud Run — stateless execution layer for repeatable assessment workloads
- Cloud Scheduler — scheduled assessment cycles for recurring cost visibility
- Terraform — modular provisioning of assessment infrastructure and supporting cloud resources
- Reporting pipeline — structured transformation of collected facts into findings and recommendations
What this project delivers:
- Multi-project cloud cost and resource assessment
- Deterministic fact-to-finding evaluation workflow
- Prioritized optimization recommendations with supporting evidence
- Repeatable assessment structure suitable for client delivery and governance reviews
Open-source note:
Selected components, architecture documentation, and implementation patterns are published publicly while protected commercial logic remains separate. The earlier Cloud Optimization Engine represents the open-source foundation from which this platform evolved.
Engineering Approach
All flagship projects follow consistent engineering principles:
- Architecture-first design with clear system boundaries and responsibilities
- Integrated security controls across identity, infrastructure, AI agents, and delivery workflows
- Cost-aware infrastructure decisions and resource modeling
- Structured documentation covering architecture, deployment, security, operations, and cost analysis
- Troubleshooting guides and operational scenarios captured as part of implementation
- Modular infrastructure and reusable configuration patterns
- Infrastructure as Code as the default provisioning model
Supporting Work
A curated collection of supporting implementations, AI infrastructure projects, experiments, and reusable engineering patterns complementing the flagship platform projects.
Includes:
- AI infrastructure and distributed agent systems
- CI/CD implementations (GitHub Actions, Cloud Build, Jenkins)
- Multi-cloud infrastructure examples (AWS, Azure, GCP)
- Kubernetes patterns (Ingress, Istio, Kustomize, OpenShift)
- GitOps workflows (Argo CD, Helm)
- Observability setups (OpenTelemetry, Prometheus, Grafana, Datadog)
- Infrastructure modules (Terraform, Ansible)
- Automation scripts (Python, Bash)
