
Hi! I'm
Matheus Senas de Cristo
Full Stack Developer · Natal, RN
Currently at Paciente 360 and previously at Solução Tributos and MatrículaAI · IMD/UFRN.
I build backends that handle volume, React interfaces and generative AI products, from RAG agents to interactive avatars. NestJS, React, TypeScript, Generative AI
20 days → 1h
to import 5 years of tax data from 167 municipalities
-40%
in infrastructure costs with a multi-tenant architecture
1,000+
simultaneous participants in real-time surveys
About me:
I started coding by building the backend of an academic project at UFRN, and ever since I've enjoyed working where the problem is big and the impact is visible: turning a 20-day manual process into a 1-hour routine, or getting a legacy system running on students' phones.
Today I work in health education, building generative AI products that help teachers design clinical cases and let students practice with interactive avatars. I also use AI tools in my day-to-day development, from Claude Code to my own skills and agents.

My experience:
Paciente 360
Dec. 2025 – Present
Full Stack Developer
Dec. 2025 – Present
- ▸Built an orchestrator of specialized AI agents that helps teachers create clinical cases, with RAG over validated protocols and materials.
- ▸Created a clinical case simulation product with AI avatars (HeyGen, ElevenLabs) and real-time feedback, cutting costs by removing the need for recordings with real people.
- ▸Extended a legacy fixed-answer module to accept free-form responses evaluated by AI, showing each student's difficulties in detail.
- ▸Migrated the legacy case player to NestJS and React, bringing cases to mobile devices as well.
- ▸Helped rebuild real-time surveys with Redis, validated for over 1,000 simultaneous participants.
- ▸Fixed vulnerabilities with RBAC and wrote Jest tests to protect critical business rules.
Solução Tributos
Feb. 2024 – Dec. 2025
Full Stack Developer
Feb. 2024 – Dec. 2025
- ▸Automated the import of five years of tax data from the 167 municipalities of RN via the SEFAZ-RN API: from ~20 manual days to ~1 hour, with fewer errors.
- ▸Optimized processing of large invoice batches with idempotency, using k6 load tests to find bottlenecks in the workers.
- ▸Implemented queues with BullMQ and Redis, with retries and automatic recovery, and tuned queries and indexes over millions of records.
- ▸Designed a multi-tenant architecture with data isolation, contributing to a 40% reduction in infrastructure costs.
- ▸Built React interfaces that help auditors spot divergences in municipal transfers (ICMS, FUNDEB, royalties and FPM).
MatrículaAI · IMD/UFRN
Jul. 2023 – Feb. 2024
Junior Developer
Jul. 2023 – Feb. 2024
- ▸Designed the architecture and built the backend from scratch, in a team of three, to support enrollment planning at UFRN.
- ▸Implemented enrollment simulation following the official rules, validating schedule conflicts and prerequisites, with REST APIs and JUnit tests.
- ▸Integrated academic data and SIGAA account login via OAuth2. Project presented at VI Tecnicamente (STI/UFRN).
What I work with
My everyday tools, from the database to AI.
Backend
Frontend
Data & Infra
Architecture & Quality
AI
Education:
Bachelor's Degree in Information Technology
Federal University of Rio Grande do Norte
Mar. 2024 – Dec. 2026 (expected)
Languages:
- PortugueseNative
- EnglishC1
Let's talk?
Open to new opportunities and to chatting about backend, data and generative AI.