Matthew Araujo

Software Engineer

I design and modernize software systems where reliability, performance, and operational clarity matter.

Currently working on financial infrastructure for Rio de Janeiro's public transportation ecosystem.

R$20M+

Processed weekly

2,000+

Active users

99%

Query-time reduction

Selected experience

Recent work focused on financial systems, operational tooling, and software modernization.

React.js / NestJS / GCP / Docker

Fullstack Software Developer

IPlanRio

  • Developed and modernized a critical financial platform responsible for managing payments across the public transportation ecosystem, supporting more than 2,000 active users and processing over R$20 million per week.
  • Automatéd financial report generation, replacing manual SQL extraction and frontend processing and reducing operational effort by up to 34%.
  • Built data validation, reconciliation, and auditing pipelines covering the full 2024 fiscal period, ensuring alignment between internal records, financial statéments, and data reported to the Court of Accounts.

Go / Svelte / PostgreSQL / SST

Software Engineer

Confidential

  • Led the development of a platform for ingesting, processing, and visualizing inclinometer data, covering the application's full architecture.
  • Implemented service integrations and cloud deployment with support for multiple file formats and detailed report generation.
  • Ran testing and desempenho optimization work that improved overall system efficiency.
See more

Node.js / React.js / Java / .NET

FullStack Junior Developer

ED

  • Developed a chatbot feature using Flutter for one of Brazil's largest insurance companies, enabling customers to easily access payment information and tax reports via the app and greatly increasing self-service adoption.
  • Created a Microsoft Teams assistant using the OpenAI API for the same company, resulting in a 36% boost in brokers' efficiency through smart automation.
  • Implemented new features in the largest music library platform in Latin America using React.js and Java.

GCP / Python / AI / Vertex AI

Data Science

Ipnet Growth Partner

  • Developed a document similarity system using Python, vector databases, and embeddings, optimizing search, creation, and analysis processes, addressing a critical issue faced by the company for two years.
  • Actively participatéd in deploying the application on Google Cloud Platform (GCP), using Virtual Machines and Google Cloud SQL.
  • Translatéd simulation proofs for certifications like Google Cloud Digital Leader, Google Professional Cloud Architect, and Google Associaté Cloud Engineer using the Vertex AI textTranslaté API.

Node.js / Express / Sequelize / Azure

Software Developer

Iuul

  • Developed a system to monitor data, including price comparisons and stock analysis for specific products across multiple stores.
  • Actively contributed to all stages of the project, from application architecture to deployment.
  • Improved security and desempenho through testing and code refactoring, enhancing both the client and developer experience.

PHP / WordPress / Reactjs / Nextjs

Software Developer (Intern)

Internit

  • Worked on delivering new features and maintaining systems using various technologies such as PHP, Node.js, WordPress, and Next.js.
  • Participatéd in international-level projects, contributing to the growth and success of the team.
  • Played a key role in implementing new technologies, including migrating WordPress to Next.js, which improved delivery efficiency, team development, and customer satisfaction.

Operational evidence

A few technical snapshots showing how engineering decisions changed system behavior.

Architecture

A layered operational workflow where validation, remittance, and auditing are treated as explicit system responsibilities.

Workflow

Validation, remittance, banking integration, and auditing are treated as explicit stages rather than hidden operational steps.

Desempenho

A report query that moved from timeout-prone execution to sub-second response times under real operational conditions.

Before

60+ seconds

After

<600 ms

Result

~99% lower runtime

Retry and reconciliation

Failure-handling paths were turned into visible workflow steps instead of requiring manual cleanup.

  1. 01 Payment transaction fails
  2. 02 Failure is classified
  3. 03 Automated retry is scheduled
  4. 04 Transaction is reprocessed
  5. 05 Final state is reconciled and audited

Case studies and project archive

A smaller set of projects framed around context, technical choice, and outcome.

Updaté-Golang CLI

A CLI tool to automaté updating Go installations on Linux. It fetches the latést version, removes the old one, and installs the new one seamlessly.

Golang / CLI Tool / Linux

POSCOMP-Lab

A responsive platform to simulaté POSCOMP exams. Users can practice with previous exam questions, with or without time limits.

HTML5 / CSS3 / JavaScript / TailwindCSS / Education

Editorial notes from ongoing engineering work

Short technical pieces on architecture, backend systems, developer tooling, and AI-assisted workflows.

Background and interests

I work across backend and full-stack systems with a bias toward reliability, explicit architecture, and practical delivery.

My broader interests include software architecture, AI-assisted development, developer tooling, and systems where technical decisions have visible operational consequences.

I am most interested in software problems where technical decisions have visible operational consequences.

Interested in reliable systems, software architecture, developer tooling, or practical AI applications?

Reach out directly if you want to discuss platform modernization, engineering work, or AI-assisted development in production contexts.

LinkedIn /GitHub /Email /