Arthur Oliveira

As a Software and DevOps Engineer, I took part in the digital transformation of different areas of the company, ranging from tax-related products to corporate messaging. My focus has always been on creating scalable solutions that address real customer bottlenecks and deliver direct business impact. Key contributions:

Menufaz (Digital Menu)

• When I joined, the system faced severe issues in the ordering flow. I refactored the backend (PHP 7 + MySQL) and the frontend in JS, creating a more reliable B2B experience for restaurants and bars.
• Integrated automatic payments (Asaas), accelerating the billing process by 80%.
• Integrated Google Maps for georeferencing restaurants and bars to automatically define delivery fees, resolving friction and speeding up the ordering process by 60%.
• Redesigned the infrastructure with staging and server migration, reducing downtime by 90%, cutting server costs by 67%, and ensuring 99.9% availability.

Plataforma do Contador (Tax Management):

• Took on the challenge of transforming a heavy, paper-based, and manual tax process into a robust digital platform. Designed the full-stack architecture (React/Vite + NestJS) with a focus on performance, enabling tax calculations and integrations to run smoothly even during peak periods.
• To handle large volumes, implemented asynchronous queues with RabbitMQ and storage with S3, eliminating bottlenecks and quadrupling processing capacity, ensuring stability even with massive invoice batches.
• In addition to the web system, developed a cross-platform desktop application in Tauri that automatically synchronizes clients’ invoices. This automation replaced time-consuming manual work, reducing accounting and support teams’ workload by over 70%.
• Structured end-to-end testing (Jest) and a complete deployment pipeline on Linux (Nginx), which increased delivery confidence and reduced production errors.

ConectaFaz (Messaging):

• The system was already in production but suffered from critical issues: delayed message delivery, unexpected connection drops, and request limit problems.
• Worked on the backend (Node.js + Express) and frontend (React + Material-UI), fixing structural bugs and ensuring greater reliability in message delivery.
• Restructured real-time communication with Socket.io, which stabilized the platform and reduced connection failures.
• Fixed and expanded the notifications module, enabling support teams to be alerted in real time about pending requests, accelerating response times.
• Implemented an internal kanban for tracking support cases, with tagging and employee filters, improving organization and traceability.
• Added usage analytics modules to monitor team and client demands (who requests the most support, which issues are recurring), providing valuable insights for decision-making.
• Result: The platform evolved from unstable and hard to maintain into a reliable corporate messaging system, supporting dozens to hundreds of daily messages with much greater operational visibility and control.

OsTicket (Customization):

• Fixed bugs and adjusted internal workflows in OsTicket, including email configuration. These improvements increased support efficiency by 20%, reduced team rework, and minimized communication friction.

Internal Bot:

• Developed an internal Python bot for monitoring and alerts via Discord. This reduced incident response time by 80%, as the team was notified in real time about critical failures before complaints reached the support team.

Beyond my professional work, I have developed large-scale personal and private projects, always focused on scraping, distributed architectures, and high-availability systems. Key achievements:

Yuppi (Fashion and Beauty Search & Discovery):

• Designed and developed a full-scale product scraping and indexing platform, combining Next.js, NestJS, PostgreSQL, and Redis, with a focus on scalability.
• Implemented an intelligent search system with Typesense and an AI-based image categorization microservice, increasing search accuracy by 30% and automating product classification.
• Optimized Python scraping with anti-DDoS techniques, reducing data collection blocks by 95%.
• Structured cascading side-effects with BullMQ and CI/CD pipelines in GitHub Actions, decreasing processing failures by 70% and bringing predictability to deliveries.

Nazumi (Discord Bot):

• Created and maintain Nazumi, a bot for anime communities that grew organically to 500,000 users across 400 servers, with 10,000 daily active users.
• Developed high-engagement features such as ranking, virtual currency, games, XP, RSS feeds, automatic roles, and notifications, increasing user engagement by 60%.
• Ensured stability for a massive user base with an architecture using TypeScript, MongoDB, and Discord.js, along with full infrastructure management.

Private Projects (Large-Scale):

• Collaborated on video streaming and media platform projects, contributing to the development of custom player systems and control dashboards.
• Worked on automating scraping across multiple platforms to collect videos, descriptions, and metadata, ensuring organized and consistent data.
• Implemented user experience improvements, such as player enhancements, increasing efficiency and user satisfaction by approximately 70%.