VP Platform Engineering

Mission. Own the platform end to end through the scale-up: what the system is and how it runs. You decide how monoliths becomes an orchestration core with clear domain boundaries, and you decide how that core runs on AWS across three availability zones with zero-downtime deploys. You are accountable for availability: the on-call rotation, an SLO per payment flow, incident command and the escalation path for the 24/7 Mission Control desk. This is a staff-level player-coach role: a third of your week is design documents, prototypes and the hardest reviews; a third is building the runtime with your team; a third is leading.

Reports to the CEO, as a peer of the Head of Engineering and the CISO. Leads the SREs, the infrastructure engineer and the Mission Control lead; the domain owners answer to you for design decisions.

What you will own: architecture

• The target architecture and its sequence: the orchestration core out of the monolith, an event backbone with outbox and idempotency, data ownership per domain, the second region.
• The design review gate: a one-page design before anything touches money movement, customer data or capacity, with decision records as the system of record.
• Standards: API design, versioning, pagination and rate limits, webhook contracts, correlation ids, security by design together with the CISO.
• Integration architecture: bank connections, routing and failover, reconciliation and settlement flows, real-time payment rails.

What you will own: platform

• Runtime on ECS: task sizing, CPU and memory limits, gunicorn and Celery topology, health checks, graceful shutdown, rolling or blue-green deploys.
• Data tier: RDS to Aurora with near-zero downtime, RDS Proxy and IAM authentication, read replicas, backups, point-in-time recovery, cross-region replication. Redis to Valkey on ElastiCache. SQS and dead-letter queue hygiene, Kinesis capacity, Lambda concurrency controls.
• Edge and network: ALB and NLB, WAF in block mode, per-merchant rate limiting, TLS, VPC design, private connectivity to providers, a multi-account landing zone.
• Infrastructure as code and delivery: Terraform modules, GitHub Actions pipelines, environment parity, drift detection, rollback.

What you will own: reliability

• The team: hire and lead SREs,
• Build the AI NOC (network operation center)
• SLOs per payment flow, error budgets, and the capacity gates that decide when volume ramps; load tests at 10x and 30x; monthly game days; the disaster recovery plan with a tested runbook.
• Incident response: incident command, the escalation policy, blameless postmortems with owners and dates, and the five pre-approved actions Mission Control can run on its own.
• Observability and cost: the Datadog APM rollout, SLO dashboards, capacity and cost visibility, Savings Plans and right-sizing.
• PCI DSS infrastructure controls and SOC 2 evidence with the CISO; architecture reviews with tier-1 merchants, acquirers and auditors.

How you will work

• Design it, build the first version with your team, hand it over with a runbook, then hold the pager like everyone else.
• One weekly architecture and operations review that you chair; every decision written and visible inside the company.
• Every production change goes through code review and a Terraform plan; no console changes.
• Ship in increments behind gates, never a big-bang rewrite.
• In year two the role splits: you keep the half where you add more value and hire the other.
• Work in Spanish and English with a distributed team across the Americas.

VP Platform Engineering: what we need from you

Must have

• 12+ years building software: at least 2 designing distributed transactional systems that moved money or handled regulated data
• Payments depth: authorization and capture, idempotency, ledgers and reconciliation, refunds and disputes, real-time rails such as SPEI, and the PCI DSS scope decisions behind them.
• Expert in Python with Django and Django REST Framework, Celery, and PostgreSQL at scale: partitioning, replication, migrations without downtime, Aurora. You know what gunicorn and Celery do under memory pressure.
• AWS in production: ECS on EC2 and Fargate or Kubernetes, Aurora or RDS, ElastiCache or Valkey, SQS, Kinesis, Lambda, DynamoDB, API Gateway, ALB and NLB, WAF, Route 53, VPC design, IAM and multi-account governance.
• Terraform at module level with remote state and plans in CI; GitHub Actions or similar; Python and Bash for automation.
• Event-driven architecture and API design for third parties: versioning, pagination, rate limiting, webhooks with retries and signatures.
• Document stores and search indexes as read models, MongoDB or equivalent; enough TypeScript on Node to set standards for a serverless estate.
• Has built an on-call rotation from scratch and run incident command for severity-one incidents; can show the postmortems.
• Linux and container debugging: out-of-memory behavior, cgroups, page cache and swap, I/O saturation, network timeouts.
• Observability as a design input: Datadog or similar, SLOs, capacity planning from production data.
• Security by design: multi-tenant authorization, input validation, secrets management; has worked inside PCI DSS or SOC 2 scope and knows what the auditor asks for.
• Has led architects or staff engineers, has hired and managed engineers, and they stayed.
• Professional Spanish and English.

Nice to have

• Took a monolith to services in production; ran at least one zero-downtime database or runtime migration.
• Multi-region active-passive on AWS that was actually tested: Aurora Global Database, DynamoDB global tables, Route 53 failover.
• AWS Solutions Architect Professional or DevOps Engineer Professional.
• Kafka, ISO 8583 or acquiring host integrations, 3-D Secure and network tokenization.
• Mexican payments regulation and rails: SPEI, CoDi, CNBV and Banxico requirements.
• Has run a 24/7 operations desk or built the escalation path for one.