Software Engineer
Grupo Boticario · Curitiba, Brazil
National-scale digital platforms with Node.js, React, and AWS — throughput, availability, and production load growth.
Node.js · React · AWS · TypeScript
👋 Hi, I'm Lucas
Today I work as a software engineer at Grupo Boticario, building platforms with Node.js, React, and AWS.
10+ years in technology: full stack, APIs, microservices, legacy modernization, and CI/CD in production environments.
Stack: Node.js · React · Next.js · TypeScript · NestJS · .NET · AWS · Docker · Kubernetes · PostgreSQL · GraphQL · CI/CD
You can find me on social networks as @olucasleitedev ⚡
Experience
AboutSoftware Engineer
Grupo Barigui · Curitiba, Brazil
Support, evolution, and modernization of corporate systems for one of Brazil's largest automotive groups.
PHP · .NET · Angular · Node.js · TypeScript · RabbitMQ · PostgreSQL · MongoDB · Jenkins · ArgoCD · Docker · AWS
Software Engineer
InvestMoney SA · Curitiba, Brazil
Financial-sector web applications with React, Next.js, NestJS, and microservices architecture.
React · Next.js · TypeScript · NestJS · Node.js · PostgreSQL · MongoDB · Docker · Kubernetes · AWS · GitHub Actions
Technical solutions in production
Microservices architecture for 100+ stores operation
Context: Sales, after-sales, and external integration flows were highly coupled, increasing deployment risk and outage windows.
- —Split domains into services with clear ownership and versioned contracts.
- —Introduced RabbitMQ/Kafka for asynchronous workloads and failure isolation.
- —Implemented idempotent consumers and retry policies with backoff for critical events.
Outcome: Lower deployment impact, fewer failures caused by synchronous dependencies, and more predictable throughput during peak traffic.
API contract standardization across squads
Context: Inconsistent endpoint semantics and unmanaged changes were breaking internal and external integrations.
- —Defined REST/GraphQL standards for error semantics, versioning, and observability.
- —Added contract validation in CI pipelines to prevent unplanned breaking changes.
- —Created concise technical documentation to accelerate onboarding and alignment.
Outcome: Less integration rework, faster onboarding, and stronger API consumption stability across systems.
Continuous delivery hardening for critical environments
Context: Manual release process was increasing operational risk and slowing down production delivery.
- —Automated CI/CD pipelines with quality gates (lint, tests, build, and security checks).
- —Containerized services and standardized rollout strategy to reduce regression risk.
- —Centralized telemetry to detect performance degradation immediately after deploy.
Outcome: Shorter lead time, safer and more frequent releases, and faster incident recovery.
Studies on the blog
View blog- Context engineering: skills and MCP — Context assembly as code — versioned skills, MCP for live systems, and governance.
- LLM agents in production — Orchestration, tools as APIs, idempotency, async workers, and observability.
- Production RAG — Hybrid search, golden-set eval, abstention, and when to skip RAG.
- Quby: RAG, text and voice agents — Multi-tenant study with pgvector, orchestrator, and Vapi — try demo.quby.com.br.
- NestJS cron on Kubernetes — Cluster CronJob instead of @Cron on every scaled API pod.
- Clean Architecture and DDD in .NET microservices — Entities, aggregates, repository, and application services with ASP.NET Core.
Selected open source
View projects- Go Notes API — Go microservices with API Gateway, Postgres per service, async audit via Kafka, and a React frontend.
- Kafka Serverless Simulator — Visual Kafka serverless simulator: partitions, consumer groups, offsets, and DLT in the browser.
- Simulador Motor API — Backend API for automotive simulation with maintainable REST contracts and modular architecture.
- gRPC Microservice Demo — Service-to-service communication example using gRPC and Protocol Buffers.