Hola, soy
Enrico
Perania
Ingeniero de Software con ~9 años de experiencia construyendo sistemas escalables en frontend, backend, cloud e IA. Expertise profundo en React, Java/Spring, Go, Python, Node.js y arquitecturas cloud-native en AWS, GCP y Azure. Historial comprobado entregando funcionalidades de alto impacto en equipos distribuidos y multiculturales.
Sobre Mí
Construyendo sistemas que escalan
Idiomas
Español
Fluido
Inglés
Fluido
Japonés
Fluido
Educación
Licenciatura en Ciencias de la Computación
Universidad Nacional de Ingeniería (Perú)
2016
Soy un Ingeniero de Software con ~9 años de experiencia construyendo aplicaciones escalables centradas en el producto. Expertise profundo en múltiples stacks: React y TypeScript en frontend; Java/Spring Boot, Go, Python, Node.js y C#/.NET en backend; y arquitecturas cloud-native en AWS, GCP y Azure.
Especializado en sistemas distribuidos, arquitecturas event-driven, integración de IA/LLM y pipelines de observabilidad. Profesional multilingüe con experiencia en equipos remotos multiculturales en LATAM, EE.UU. y Europa.
9+
Years Experience
professional
4
Global Companies
MX · US · ES · PE
3
Languages
ES · EN · JP
Core Stack
Habilidades Técnicas
Stack completo — frontend, backend, cloud e IA
Frontend
Backend
Cloud / DevOps
Datos y Mensajería
IA y Machine Learning
Observabilidad y Testing
Experiencia Laboral
Mi trayectoria profesional
- Diseñé y escalé microservicios usando Spring Boot, WebFlux y Kafka, soportando ~1M+ eventos diarios de transacciones financieras con mejora de latencia del 30%
- Construí herramientas internas potenciadas por IA usando OpenAI + pipelines RAG con Pinecone y pgvector, reduciendo la carga de soporte manual en un 50%
- Implementé búsqueda vectorial con embeddings para flujos de detección de fraude, reduciendo el tiempo de investigación en un 40%
- Lideré la modernización de arquitectura con Spring Cloud y microservicios en Kubernetes (GKE/AKS), logrando 99.95% de disponibilidad
- Establecí estándares de observabilidad con OpenTelemetry + Prometheus/Grafana, reduciendo el tiempo de respuesta a incidentes en un 30%
- Mentorié a 5+ ingenieros y lideré la entrega cross-funcional en un entorno completamente remoto
Selected Work
Proyectos
06 cases
Plataforma de Crédito con IA
El desafío
Kueski's credit decision process took 15+ minutes per application with a 34% false-negative rate. The legacy system relied entirely on manual review, creating bottlenecks at scale for Mexico's largest BNPL platform serving 2M+ users.
La solución
Designed a RAG pipeline architecture combining OpenAI GPT-4 with proprietary ML models on GCP. Built event-driven microservices with Spring Boot and Kafka processing ~1M+ daily events. Integrated vector-based search using pgvector and Pinecone for real-time policy retrieval. Deployed on GKE with Terraform-managed infrastructure. Established MLOps practices for regulated financial systems.
80%
Credit decision time reduced
99.9%
SLA achieved
35%
Cloud cost savings
Pipeline de Eventos Financieros
El desafío
Fraud investigators worked with stale batch-processed data, causing false negatives. Financial dashboards refreshed every 24–48h. Compliance risk during volatile market periods was growing.
La solución
Built reactive microservices with Spring Boot + WebFlux consuming from Kafka topics. Transaction embeddings generated asynchronously and indexed in Elasticsearch for vector-based fraud detection. Real-time dashboard updates via WebSocket hub. Observability stack: OpenTelemetry instrumentation, Prometheus metrics, Grafana dashboards, and Jaeger distributed tracing across all services.
40%
Investigation time reduced
~1M+
Daily events processed
30%
Incident response faster
Plataforma de Ejecución Distribuida
El desafío
Retool's synchronous query engine bottlenecked under load, with no tenant isolation and vertical scaling limits. The monolithic architecture caused P0 incidents every sprint and slowed new feature delivery by 3–4 weeks.
La solución
Architected an event-driven query execution system using Go worker pools, Node.js API layer, and Kafka as the event backbone. Built real-time sync pipeline for 500+ database integrations using Kafka Streams and Redis. Implemented circuit breakers with Resilience4j. Deployed on AWS EKS with Terraform. Unified auth, added GraphQL + REST APIs, and rolled out Prometheus/Grafana/ELK observability across all services.
99.99%
Uptime (3 years)
70%
P99 latency reduction
3×
Deployment frequency
Migración Monolito a Microservicios
El desafío
Factorial's monolithic Rails codebase had 45-minute deployments, no service isolation, and growing pains serving 50K+ users across Europe. Every refactor was a regression risk with zero E2E coverage.
La solución
Led the migration using Strangler Fig pattern — routing via Node.js API Gateway, decomposing domains into Spring Boot and Node.js microservices with per-service PostgreSQL databases. RabbitMQ and Azure Service Bus for async event flows (payroll, approvals). Redis caching reduced API response times by 60%. Frontend modernized with React + Redux. Mentored 3 junior engineers through the process.
91%
Deploy time reduction
€10M+
Monthly payroll processed
60%
API response time cut
Plataforma HR Escalable
El desafío
Factorial's HR modules used legacy Rails views mixed with ad-hoc React components. No shared state strategy led to stale UI after mutations, inconsistent loading states, and HR managers abandoning workflows.
La solución
Rebuilt scheduling, attendance, and employee profile backends as Spring Boot and ASP.NET Core Web APIs with clean PostgreSQL schemas. Implemented OAuth2/OIDC with PKCE for per-tenant SSO. Frontend migrated to React + Redux Toolkit with React Query for server-state sync. Built shared component library in Storybook cutting feature development time across teams.
50K+
Users served
25%
User retention increase
3×
Faster feature delivery
Modernización Plataforma SaaS
El desafío
Holded's cloud ERP platform had a legacy jQuery/AngularJS frontend with inconsistent UI, no component reuse, zero i18n support, and slow data loading. The electronic invoicing module needed to scale to thousands of SMEs across Spain.
La solución
Migrated invoicing, accounting, and inventory modules to React + TypeScript with a Material UI component library. Integrated Node.js and Vue.js services for specific product areas. Built i18n framework supporting 4 European languages. Optimized REST API integrations and introduced Storybook design system. The electronic invoicing module was adopted by 10K+ companies. PostgreSQL performance tuned with 8× improvement on key queries.
10K+
Companies adopted module
8×
PostgreSQL performance gain
35%
Load time improvement
Product screenshots © Kueski · Retool · Factorial · Holded — shown for portfolio reference
Contacto
¿Tienes un proyecto en mente? Hablemos.
Ubicación
Av. Universitaria 742, Urb. Ingeniería, San Martín de Porres, Lima, Perú
Correo
enricoperania@gmail.com
Teléfono
+1 928-268-8561
Disponibilidad
Abierto a oportunidades remotas
Currículum
Descarga o consulta mi CV completo