Skip to content

Principal Solutions Architect · Full-stack Engineer · AWS Certified

I designagentic cloud systems from data to UX

Agents, RAG, and distributed backends paired with delivery pipelines and product-shaped architecture so teams ship with confidence.

> Architecting enterprise systems that stay understandable at scale.

alajuela, costa rica · utc-6:00

[ 00 // credentials ]

Credentials grouped by issuer in the strip below. Auto-scroll pauses on hover or focus; steer with pointer left/right or arrow keys on a focused link.

All certifications

[ 01 // systems_i_design ]

Capabilities that ship as systems

Eight lenses I use in practice, from lessons learned and patterns I've seen work.

01
AI & Agentic Systems

Right outcomes

Production-ready AI beyond simple LLM calls: context, decisions, and scale, Cloud Provider Model Garden applied to real business needs.

Reliable, explainable systems, not trend-chasing demos.

  • RAG pipelines and hybrid retrieval strategies.
  • Agent orchestration with enterprise constraints in mind.
Vertex AIGeminiAzure OpenAIRAGagentsorchestration
RIGHT OUTCOMES
02
Cloud & Distributed Architecture

Provider-agnostic

Architectures across GCP, Azure, and AWS with scalability, security, and flexibility, avoiding unnecessary lock-in.

Components chosen for how they interact under real workloads.

  • Cloud Run, Pub/Sub, Service Bus, and secure networking patterns.
  • Trade-offs, failure modes, and operational impact first.
GCPAzureAWSCloud RunPub/SubService Bus
PROVIDER-AGNOSTIC
03
Backend Engineering

Predictable at scale

Domain-driven APIs, event-driven design, and async workflows in Python and Node.js, queues, webhooks, and distributed patterns.

Resilience, idempotency, and maintainability when it gets stressful.

  • Clean boundaries between services and domains.
  • Systems that behave predictably under load.
PythonNode.jsAPIseventsqueueswebhooks
PREDICTABLE AT SCALE
04
Frontend & Fullstack Thinking

Complexity, simplified

Next.js, React, Tailwind, and ShadCN, interfaces that expose powerful backends without overwhelming users.

UX is the control plane for AI: clarity where decisions happen.

  • Full-stack lens: frontend as part of the system, not an afterthought.
  • AI-driven apps with visible, trustworthy interaction.
Next.jsReactTailwindShadCNUXfull-stack
COMPLEXITY, SIMPLIFIED
05
DevOps & Platform Engineering

Ship & govern

Terraform, GitHub Actions, and container-based delivery, pipelines and internal platforms that scale with teams.

From CI/CD toward intelligent, self-improving engineering loops.

  • Automated reviews and standards enforcement where it helps.
  • Infrastructure and app delivery as a cohesive platform story.
TerraformGitHub ActionsDockerCI/CDplatform
SHIP & GOVERN
06
Data & RAG Systems

Ground truth

Pipelines for AI workloads: ingestion, transformation, and retrieval at scale for structured and unstructured sources.

Agents act on evidence, embeddings, search, and filters working together.

  • Robust retrieval: embeddings + search + structured filters.
  • Data shaped for downstream AI, not one-off exports.
embeddingssearchingestionETLlakehouseRAG
GROUND TRUTH
07
Product & Pre-Sales Architecture

Options, not ultimatums

Architecture with product sense: cost, scalability, time-to-market, and team maturity in every decision.

Stakeholders get clear trade-offs, not a single “silver bullet” deck.

  • Reusable frameworks and decision engines for solution design.
  • Pre-sales acceleration without hand-wavy boxes.
pre-salestrade-offscostroadmapstakeholders
OPTIONS, NOT ULTIMATUMS
08
Engineering Mindset

Built to evolve

Systems thinking: separation of concerns, boundaries, and long-term evolution, data placement, service interaction, scale, and failure.

Features age; architectures that adapt stay valuable.

  • Explicit boundaries and contracts between capabilities.
  • Design for change, not just the next sprint.
systemsboundariesmaintainabilityfailure modes
BUILT TO EVOLVE

[ 02 // about_me ]

There's more than a capability list behind this site.

On About, I walk through who Josue is, how I think in practice, and the architecture_flow I use from product framing to measurable, safe systems in production.

[ 03 // from_the_wiki ]

Latest from the wiki

Long-form notes on AI-assisted engineering, fundamentals that still matter, and patterns that keep delivery predictable.

Browse all articles

[ wiki // ]

Higher-order functions, composition, currying, pure functions, immutability, and avoiding shared state in JavaScript.
javascriptfunctional-programmingfpreact

[ wiki // ]

Interfaces as contracts, polymorphism, extensibility, testing, and why the extra files are worth it.
oopinterfacescsharpsolid

[ wiki // ]

A walk through Redux fundamentals: store, actions, reducers, middlewares, slices, combining reducers, and subscriptions.
reduxjavascriptreactstate-management

[ 04 // next_step ]

Let's build something solid.

Whether you want to compare notes on cloud architecture, discuss an AI-assisted workflow, or explore a consulting engagement, I'd love to hear from you.