<metaname="description"content="Production cloud infrastructure engineering — Kubernetes clusters, AI/ML platforms, GitOps pipelines, and database fleets. I build and manage infrastructure so your team ships features.">
<h1>I build infrastructure<br>that <spanclass="accent-word">stays out of the way.</span></h1>
<pclass="hero-description">
Solo infrastructure engineer for teams who'd rather ship features than debug cluster networking. <strong>Production clusters, AI inference platforms, GitOps pipelines, hardened databases.</strong> Everything versioned, everything automated.
</p>
<divclass="hero-cta">
<ahref="#contact"class="btn-primary">Get in touch</a>
<h2class="section-title">Technologies I work with daily</h2>
<olclass="stack-list">
<liclass="stack-item">
<spanclass="stack-num">01</span>
<divclass="stack-content">
<h3>Orchestration & Clusters</h3>
<p>Multi-region Kubernetes, auto-provisioning node pools, workload identity federation. Clusters designed to run reliably from day one.</p>
<divclass="tags">
<spanclass="tag">Kubernetes</span>
<spanclass="tag">GKE Autopilot</span>
<spanclass="tag">kubeadm</span>
<spanclass="tag">Multi-cluster</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">02</span>
<divclass="stack-content">
<h3>CI/CD & GitOps</h3>
<p>Every commit propagates to cluster — auditable, repeatable, fast. Automated dependency updates, zero-downtime rollouts, instant rollback if something breaks.</p>
<divclass="tags">
<spanclass="tag">GitHub Actions</span>
<spanclass="tag">Flux CD</span>
<spanclass="tag">Zero-downtime</span>
<spanclass="tag">Auto-rollout</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">03</span>
<divclass="stack-content">
<h3>AI/ML Infrastructure</h3>
<p>GPU-accelerated serving with scale-to-zero autoscaling, model routing, multi-tenant isolation, and cost tracking. From lightweight dispatch to heavy inference.</p>
<divclass="tags">
<spanclass="tag">GPU Serving</span>
<spanclass="tag">vLLM</span>
<spanclass="tag">Model Routing</span>
<spanclass="tag">Scale-to-Zero</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">04</span>
<divclass="stack-content">
<h3>Data & Storage</h3>
<p>High-availability PostgreSQL with automated backups, point-in-time recovery, and connection pooling. Time-series databases and message queues for event-driven architectures.</p>
<divclass="tags">
<spanclass="tag">PostgreSQL</span>
<spanclass="tag">CNPG</span>
<spanclass="tag">PgBouncer</span>
<spanclass="tag">Time-series</span>
<spanclass="tag">Message Queues</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">05</span>
<divclass="stack-content">
<h3>Networking & Security</h3>
<p>Network segmentation, mTLS, encrypted secrets with KMS unsealing, and CIS benchmark compliance. Defense in depth applied by default.</p>
<divclass="tags">
<spanclass="tag">Cilium</span>
<spanclass="tag">Gateway API</span>
<spanclass="tag">mTLS</span>
<spanclass="tag">Encrypted Secrets</span>
<spanclass="tag">KMS</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">06</span>
<divclass="stack-content">
<h3>Observability</h3>
<p>Prometheus metrics, Grafana dashboards, structured logging, and alerting pipelines that notify before you notice. Full traceability from symptom to root cause.</p>
<divclass="tags">
<spanclass="tag">Prometheus</span>
<spanclass="tag">Grafana</span>
<spanclass="tag">Structured Logging</span>
<spanclass="tag">Alerting</span>
</div>
</div>
</li>
<liclass="stack-item">
<spanclass="stack-num">07</span>
<divclass="stack-content">
<h3>Infrastructure as Code</h3>
<p>Declarative configurations with strict separation between provisioning and application manifests. Two-repo discipline — infrastructure and apps never mix.</p>
Full autonomous pipeline: scrapes news sources continuously, runs LLM-based analysis on GPU, and pushes trading signals to messaging platforms in real time. Monitors financial markets with algorithmic strategies.
<spanclass="work-name">Multi-agent AI orchestration platform</span>
<spanclass="work-env">production</span>
</div>
<pclass="work-desc">Specialist agent teams coordinated through kanban-driven task decomposition. Persistent memory, autonomous coding, research, and operations — infrastructure managed by the very agents it runs.</p>
<divclass="work-meta">
<span>multi-agent</span>
<span>kanban-driven</span>
<span>self-operating</span>
</div>
</div>
<divclass="work-card"style="cursor: default;">
<divclass="work-card-header">
<spanclass="work-status"></span>
<spanclass="work-name">Real-time financial data infrastructure</span>
<spanclass="work-env">production</span>
</div>
<pclass="work-desc">Market data feeds, whale tracking, position scanners, and trading signal pipelines. Low-latency infrastructure running continuously on dedicated clusters with high-availability databases.</p>
<pclass="work-desc">Production clusters, high-availability databases, and GitOps deployments across multiple workloads. Terraform provisioning, automated rollouts, and continuous monitoring.</p>
<divclass="work-meta">
<span>multi-cluster</span>
<span>GitOps</span>
<span>100+ day uptime</span>
</div>
</div>
<divclass="work-card"style="cursor: default;">
<divclass="work-card-header">
<spanclass="work-status"></span>
<spanclass="work-name">Private AI inference platform</span>
<spanclass="work-env">production</span>
</div>
<pclass="work-desc">GPU-accelerated serving with autoscaling, model routing, and cost tracking. Fully isolated environments with multi-tenant support and scale-to-zero for cost efficiency.</p>
<h3>End-to-end infrastructure design and management</h3>
<p>Production clusters designed from scratch, provisioned with IaC, and managed with GitOps. Automated builds, zero-downtime updates, encrypted secrets, and monitoring that alerts before you notice. I operate what I build.</p>
</div>
</li>
<liclass="svc-item">
<spanclass="svc-num">02</span>
<divclass="svc-content">
<h3>AI/ML platform deployment and scaling</h3>
<p>GPU-accelerated inference platforms with autoscaling, model routing, multi-tenant isolation, and cost controls. From prototype to production — your models, your data, fully isolated.</p>
</div>
</li>
<liclass="svc-item">
<spanclass="svc-num">03</span>
<divclass="svc-content">
<h3>Autonomous agent system architecture</h3>
<p>Multi-agent systems with specialist teams, shared orchestration, persistent memory, and versioned operations. Agents that plan, research, code, review, and execute — with full traceability.</p>
</div>
</li>
<liclass="svc-item">
<spanclass="svc-num">04</span>
<divclass="svc-content">
<h3>CI/CD and GitOps implementation</h3>
<p>Automated pipelines from commit to cluster. Dependency updates, zero-downtime rollouts, instant rollback, and auditable history. No click-ops, no manual deployments.</p>
</div>
</li>
<liclass="svc-item">
<spanclass="svc-num">05</span>
<divclass="svc-content">
<h3>Security audits and hardening</h3>
<p>Network segmentation, non-root containers, encrypted secrets, mTLS, CIS benchmarks, and vulnerability scanning. Defense in depth applied systematically, not retrofitted after an incident.</p>
<p>Whether you need a production cluster designed from scratch, an existing setup hardened for security, or an autonomous intelligence system built for your domain — reach out. I work remotely and communicate in code reviews and commit messages.</p>
