Registration, QR Code badges, entry control, badge scanning by exhibitors to capture leads, and attendance per talk. The organizers see how many people are in the building in real time.
Mathias Novais da Silva
Software Engineer · Development, Automation, Integrations and Applied AI
- BARRA OLÍMPICA, RIO DE JANEIRO · BRAZIL
- WHATSAPP · (21) 99036-7920
- LINKEDIN · linkedin.com/in/mathiasnovais
- GITHUB · github.com/mathiasnovais
- Software Engineer · School network in Rio de Janeiro · 2025 · present
- Postgraduate Program in Artificial Intelligence and Data Science for Leaders · Mackenzie Presbyterian University
- Bachelor's in Computer Science · Estácio de Sá University (UNESA)
- English · Portuguese
- 11 projects in the portfolio, 7 systems in production with their real screens on this site.
Professional summary
Software Engineer with a Bachelor's degree in Computer Science, working on internal systems development, process automation and integration between corporate platforms. Systems in production include the university fair management platform, with more than 600 simultaneous users, and the performance review platform used across the whole school network, with cascading dashboards by team and by role. Daily work with JavaScript, Node.js, PostgreSQL, REST APIs, n8n and TOTVS RM, plus generative AI applied to data analysis.
Work experience
- Development and maintenance of internal systems for customer service, operational automation and process digitization, from requirements gathering with each department to delivery in production.
- Integrations between n8n, TOTVS RM, REST APIs and legacy systems, building end-to-end automated flows with a focus on reliability.
- Building interfaces in HTML, CSS and JavaScript and data routines in PostgreSQL and advanced SQL for queries, reports and automations across departments.
- Contributed to the corporate intranet, which centralizes internal tickets, purchase requests, documents, the policy manual and the institutional library.
- Automation of the enrollment center funnel, with lead qualification, scheduling integrated with the corporate calendar, automatic notifications and stage tracking.
- Generative AI solutions applied to the business, including a customer service chatbot and prompt engineering for data analysis.
Projects
A form with conditional logic that turns into a dashboard for each person, feedback meetings that generate minutes ready to sign, cascading access by manager and a leadership panel filtered by role. Indicators ready for AI to read.
A dashboard that reads the logs of every PostgreSQL database in the organization with a single token and shows requests, latency, errors and the origin of each access on a 3D globe. In production on the network's real data, and open here for you to try live.
Online registration, exam scores entered like a spreadsheet, automatic ranking, interviews booked by families on their phones and enrollment with documents. All through single-use links, with authorization enforced in the database.
Public five-step application, written exam with QR attendance, scores per subject, résumé review, invitations with single-use links and a booking page for candidates. Each director sees only their own school level.
The supervisor takes the photo on a phone, the browser adjusts size and quality until it meets the required standard, and an n8n workflow saves it to Drive, in the class folder, with the right file name. The coordinators see who is missing in each class.
Hundreds of teaching materials under review and rewrite, tracked by a dashboard that n8n builds by reading the spreadsheet on every request. Leadership sees what is at risk, by school level, track and stage.
A single page with countdown, photo modules, manifesto, step-by-step method, offer and FAQ, plus a built-in dashboard that shows how far each visit scrolled and where it clicked. Copy and layout designed for visitors coming from Instagram.
Internal tickets, purchase requests, documents, policy manual and institutional library in one portal. I took part in development, from interfaces to data routines.
A lead comes in, gets qualified, the visit goes on the corporate calendar, the family gets notified and the stage is recorded. Everything flows between n8n, TOTVS RM and legacy systems.
An assistant that answers families using operational data and, on the other side, structured prompts that turn the knowledge base into readable analyses and indicators, no report needed.