Staff Software Engineer ·
Designing, scaling, and operating cloud-native, event-driven platforms that process 10M+ events per day with strong guarantees around reliability, security, and observability.
$ whoami
staff-engineer · platform & distributed-systems
$ uptime --career
10+ years in production
$ kubectl top platform
events/day: 10M+ availability: 99.9%
$ stack --core
python · kafka · rabbitmq · kubernetes
$ ▊
I'm a Staff Software Engineer based in Barcelona with over 10 years of experience building and operating cloud-native, event-driven platforms in production. My work sits at the intersection of platform engineering, distributed systems design, and cross-team technical leadership.
Currently at Telavox AB, I own the end-to-end architecture of a real-time analytics platform that processes 10M+ telephony events per day — from initial design to production reliability. Previously I led IoT platform development at Amper / NVISION, and worked on mission-critical systems for airlines and enterprise clients across Spain.
I care deeply about reliability, observability, and system correctness. I enjoy mentoring engineers, driving architecture reviews, and making complex distributed systems understandable, maintainable, and built to last.
Areas where I bring deep, hands-on expertise — from design decisions to production ownership.
Designing high-throughput, fault-tolerant event pipelines with RabbitMQ and Kafka. Deep experience with Event Sourcing for auditability, temporal queries, and incident debugging at production scale.
End-to-end ownership of Kubernetes-based platforms with Helm, horizontal autoscaling, zero-downtime deployments, and multi-tenant isolation on Azure and Google Cloud Platform.
Applying DDD, CQRS, and Bounded Contexts to build systems that accurately model complex business domains while remaining evolvable, testable, and maintainable over time.
Designing for failure: idempotency, backpressure, graceful degradation, and circuit breakers. Observability stacks with Prometheus and Grafana enabling SLO-driven operations and proactive incident response.
Building internal platforms and standardized engineering practices — CI/CD pipelines, shared infrastructure patterns, and developer tooling that enable multiple teams to ship reliably and consistently.
Acting as technical authority across teams: leading architecture reviews, mentoring engineers, making complex trade-offs explicit, and creating alignment on system-level decisions that span organizational boundaries.
Telavox AB
Amper Sensing & Control / NVISION
Volotea · Vueling · Plain Concepts (Lidl Plus) · Wolters Kluwer · Indra BPO
Worked on mission-critical systems in aeronautical and enterprise environments, focusing on distributed architectures, legacy modernization, migrations to .NET Core, and adoption of Domain-Driven Design (DDD) and CQRS patterns.
Multi-tenant, real-time analytics platform built from scratch to power embedded, customer-facing dashboards inside Telavox's products. Designed as a deliberate alternative to Grafana, Cube.js, and TrinoDB — none combined server-side tenant isolation, change-aware live data over WebSocket, and clean embeddability.
End-to-end event-driven analytics platform for large-scale telephony systems. Processes 10M+ events per day (calls, queues, IVR) with strong guarantees around reliability and latency. Built on Python microservices with Kubernetes autoscaling, RabbitMQ for high-throughput messaging, and Event Sourcing for full auditability and incident traceability.
Multi-tenant cloud platform for monitoring solar plants, EV charging stations, and battery storage systems. Built as part of the EU Horizon 2020 research programme. Delivered real-time energy telemetry, analytics dashboards, and alerting for industrial-scale renewable energy assets across multiple tenants.
Asset geolocation and fleet management platform with geofencing and real-time alerting. Built for industrial environments with heterogeneous tracking devices and protocols. Designed for high availability and low-latency location updates at scale, with configurable alert rules and multi-tenant data isolation.
A skill factory and shared library for AI coding agents: captures how a team does things — processes, conventions, roles — once, in a format Claude Code reads and applies consistently across every project. Ships 13 SDLC expert personas (architect, developer, QA, security…), a meta-skill for authoring new skills, a versioned memory ledger that survives across sessions, and an always-on harness mode with human-in-the-loop gates and deterministic validators. Distributed as a Claude Code plugin marketplace, with generated integrations for Cursor, IntelliJ, and OpenAI Codex.
AI-native automated penetration testing engine that orchestrates 70+ security tools across 12 MITRE ATT&CK phases. Built as a deliberate alternative to manual red team workflows — combining a LangGraph state machine for phase orchestration, Kali Linux workers dispatched via RabbitMQ, and an optional cognitive "brain mode" with ReAct reasoning and ChromaDB RAG for hypothesis-driven exploitation. Designed for multi-tenant deployments with per-audit Elasticsearch indexing and a live React dashboard for real-time findings and ATT&CK heatmap coverage.
Open-source SOAR platform for SOC teams. Ingests multi-protocol telemetry (SNMP, SSH, WinRM, REST, NETCONF, syslog, AWS CloudTrail, Azure Monitor), normalizes it to ECS, and indexes it in Elasticsearch. Bundles threat intelligence (MITRE ATT&CK, IOCs, IP reputation), automated triage with L1-L3 / P1-P4 scoring, and a remediation engine that generates CWE-mapped code fixes and containment playbooks — gated by tech-lead approval. Hybrid AI log analysis pairs a local LLM with Anthropic / OpenAI models.
Open to Staff / Principal / Lead Engineer roles in distributed systems, platform engineering, and real-time data. Fully remote or based in Barcelona.