Gabriel Ferreira

- Migrated core search from PostgreSQL to Elasticsearch via flagged rollout with parity validation; decommissioned legacy path.
- Built self-healing resilience mechanisms (watchdogs, auto-heal jobs, lock recovery) for production AI infra at scale.
- Built CI-gated eval/safety framework for an LLM product, driving fixes to key failure patterns.
- Shaped team on-call, review, and AI-dev practices.

- Led AI product initiatives, translating ambiguous education problems into production systems via rapid experimentation.
- Owned 0→1 dev of an LLM-powered assistant for curriculum-aligned, differentiated assignments.
- Designed a production RAG architecture (multi-stage retrieval, evals, guardrails).
- Cut teacher planning effort 85%, scaling to 250+ teachers, 2K+ students, 300+ assignments/day.

- De-risked new AI capabilities by owning validation through rapid experimentation and feature-flag driven releases.
- Cut LLM-agent pipeline latency by 84% through orchestration and retrieval optimizations, including scaling MongoDB and Milvus-based retrieval systems for high-throughput workloads.

- Architected data pipelines/analytics systems processing 1M+ events/day, enabling revenue decisions across web, mobile, and geospatial channels.
- Defined system requirements in high-ambiguity spaces, iterating PoCs into production-ready solutions.
- Delivered dashboards shaping campaign optimization and budget allocation.
- Automated admin ops, saving 20+ hrs/week of manual reporting.

- Strengthened security and reliability of an enterprise-grade platform for managing and sharing compute resources as a core contributor to the team that designed, developed, and secured it. Flagship system: HP Anyware.
- Demonstrated production engineering discipline by owning security compliance workflows within CI/CD pipelines and improving system reliability through automation and tooling.

- Specialized in supply chain security, finding vulnerabilities in ecosystems like npm and the Linux kernel via custom analysis tools.
- Analyzed ~1.5M npm packages, contributing to ICSE-published research on secure dependency management.
- Strengthened research output (papers, talks, grants) and mentored students in research fundamentals

- Modernized university-wide ERP systems as a core engineering contributor.
- Built scalable, maintainable Java backend services and REST APIs while maintaining SLA targets.
- Co-owned a training initiative that raised engineering standards and knowledge-sharing across teams.

- Modernized legacy CRM, GIS, and payment systems and integrations as a core contributor (and Scrum Master for some sprints) to an engineering team within a startup embedded in a telecommunications company.
- Enabled a modern, highly-concurrent web-based portal for Telecom agents by planning, designing, and implementing a microservices platform. Flagship system: Portal de Vendas.

- Modernized enterprise VoIP infrastructure as a core contributor to an engineering team at a telecommunications company.
- Delivered user-facing software systems using Java technologies, J2EE, Spring Framework, Struts 2 Framework, and design patterns.
- Enabled Call Detail Record (CDR) collection across open-source code in OpenSIPS and Asterisk by building custom plugins in C.

Activities and societies: Communication, Python, Machine Learning, Presentation Skills, JavaScript, Scala, C (Programming Language), Mentoring, Software Architecture, Algorithms, Scientific Writing, Node.js

Pursued a Doctor of Philosophy in Software Engineering. Focused on advanced research on Software Security and mixed-methods research approaches.

Activities and societies: Communication, Presentation Skills, C (Programming Language), Mentoring, Algorithms, Scientific Writing

Completed a Master's degree in Computer Science. The dissertation focused on the use of Feature-oriented Programming for Evolving Software Product Lines.

Activities and societies: Undergraduate teaching assistant: Introduction to Computer Science. Undergraduate Research Opportunities Program student: Data mining, Tree patterns, XSLT, XQuery. UFUGames - Game development study group - Technologies/Tools: C++, Ogre3D, SDL, OpenAL. Skills: C (Programming Language), Java.

Earned a Bachelor's degree in Computer Science. Served as an undergraduate teaching assistant and participated in the Undergraduate Research Opportunities Program focusing on data mining. Was also involved in the UFUGames development study group.