WebSocket connection sessions/day
realtime deliveryTECH LEAD · INFRASTRUCTURE & REALTIME ARCHITECT
Systems that survive load, failure, and time.
I design and lead production platforms where cloud infrastructure, realtime communication, backend architecture, and operational reliability meet.
HTTP requests/day
multi-tenant SaaSMigration downtime
AWS → GCPYears engineering
build → operate → leadSELECTED SYSTEMS
Production work, not a technology inventory.
Private systems described through their engineering constraints, decisions, and measurable outcomes—without exposing customer code.
Realtime delivery that scales past a single region.
Architected WebSocket delivery across Cloudflare Durable Objects and GKE/Soketi, with Redis-backed horizontal scaling, tenant isolation, health-aware capacity, and independent release paths.
A multi-tenant platform built for sustained production load.
Led architecture across Laravel Octane, Cloud Run, GKE, Cloud SQL, queues, search, identity, observability, and tenant-aware AI services—while keeping product delivery moving.
Cloud migration treated as a controlled systems event.
Coordinated application, database, queues, DNS, and infrastructure through an AWS-to-GCP cutover, backed by infrastructure as code, GitOps delivery, observability, and a rehearsed rollback boundary.
HOW I SEE THE SYSTEM
Reliability is a property of the whole path.
Architecture starts at the user request and continues through identity, application code, data, delivery, infrastructure, telemetry, and recovery.
PUBLIC ENGINEERING
Inspectable proof.
Recent production work is private. These public references make the same engineering judgment visible.
Spring Payments Platform
An auditable payment reference: idempotent transfers, immutable double-entry ledger, optimistic concurrency, JWT scopes, signed events, transactional outbox, PostgreSQL, and Testcontainers.
WhatsApp API for Laravel
A practical Laravel integration package and one of the strongest existing public repositories in the profile.
Realtime & Platform Blueprints
Runnable reference architectures for realtime delivery, GKE/GitOps operations, observability, capacity, and failure recovery.
OPERATING PRINCIPLES
The standard I hold the work to.
Make failure explicit and recovery boring.
Observability belongs in the design, not the incident.
Remove operational work from product teams.
Treat performance, security, and cost as first-order constraints.
THE NEXT SYSTEM
Give me a hard enough problem.
I’m most useful where scale, reliability, architecture, and delivery have to work together.