Jorge García
Staff Software Engineer
Summary
Staff Software Engineer with 12+ years designing and scaling resilient, large-scale platforms across distributed systems and AI-integrated architectures. Led zero-downtime migrations from legacy monoliths to microfrontends and event-driven services. Diagnosed and resolved a throughput ceiling in distributed task queues, delivering an 8x improvement from 940 to 7,500 tasks per minute through benchmark-validated optimization. Build RAG and multi-agent systems with production engineering discipline focused on orchestration, evaluation, and observability.Technical Skills
Cloud & Platform Engineering
GCP, Kubernetes, Docker, Terraform, CI/CD, PostgreSQL, Redis
Backend & Systems Design
Node.js, Python, Go, Design Patterns, IoT Systems, Microservices
Frontend & Full-Stack Development
React, TypeScript, Next.js, State Management, API Design
AI & Distributed Systems
Multi-Agent Orchestration, RAG Pipelines, LLM Infrastructure, Event-Driven Systems, Micro-Frontends, Task Queues
Experience
Staff Software Engineer
- Led the zero-downtime migration of a frontend product from the legacy monolith to the new microfrontends architecture, executed incrementally without disrupting users.
- Designed core fulfillment workflows in Go, with schema-driven validation and graph-based traversal to guarantee data integrity across the order lifecycle.
- Led an event-driven backend integrating external accountant partners, unblocking the launch of the company's tax-related product line.
- Designed and built multiple third-party integrations with partner-specific rate limiting, timeouts, structured error handling, and secrets management.
- Co-created and maintain shared core packages and architectural patterns adopted across multiple teams, driving consistency in how services are built.
- Supported enterprise API initiatives by designing secure, public-facing endpoints that exposed internal capabilities while enforcing business constraints.
Senior Software Engineer
- Designed scalable backend features using distributed system patterns (eventual consistency, backward-compatible schema evolution) so the platform could evolve without breaking existing consumers.
- Diagnosed the root cause of a throughput ceiling in distributed task queues; a surgical configuration change, validated via benchmarks, delivered an 8× throughput improvement (~940 → ~7,500 tasks/min per queue).
- Implemented end-to-end observability with Cloud Trace, Cloud Logging, and Error Reporting, shortening incident detection through automated alerting.
Tech Lead / Digital Solutions Manager
- Led and mentored a small engineering team across concurrent client engagements, owning feature delivery and growth conversations.
- Translated business requirements into technical roadmaps and delivery plans across multiple client projects.
- Provided architectural guidance across client solutions, ensuring consistency, security, and maintainability.
- Migrated client workloads from dedicated servers to serverless platforms (Cloud Functions, Cloud Run), reducing infrastructure cost and operational overhead.
- Automated Node.js CI/CD pipelines on Cloud Build, removing manual release steps and improving deployment reliability.
- Established shared GitHub conventions and code standards adopted across teams to improve consistency and collaboration.
- Built and managed partnerships with external development providers, expanding delivery capacity while keeping quality commitments.
Full-stack Engineer / Consultant
- Built a real-time IoT platform tracking 150+ beacon devices with AWS Kinesis and MongoDB, powering live analytics dashboards for distribution centers.
- Introduced serverless and modern web technologies (Node.js, Firebase) that became foundational standards for subsequent client projects.
Technical Highlights
Deal Intelligence Platform
AI Reasoning & Strategic Intelligence
Multimodal RAG over sales calls (Whisper) and financial documents (vision). 5 specialized agents orchestrated with LangGraph; hybrid retrieval (BM25 + embeddings); evaluation with LangSmith..
Tech: Python, LangChain, LLMs, GraphRAG, Whisper + Vision
ConectaPelu2 — Animal Welfare Ecosystem
Full-Stack Product Architecture
Digital ecosystem connecting animal shelters and adopters. Monolithic Next.js app on PostgreSQL/Drizzle with pet management, secure auth, and audit trail; containerized and deployed via GitHub Actions to Kubernetes..
Tech: Next.js 15, Drizzle ORM, PostgreSQL, Docker, Kubernetes