B.S. Computer Science — Florida International University

Minor in Mathematical Sciences

SWE Intern @ Genuine Labs  |  Embedded SWE Intern @ PDS  |  INIT Build Team Lead

Backend-first software engineer with full-stack range and hands-on AI/ML integration experience. Currently building LLM evaluation infrastructure at Genuine Labs and designing an over-the-air update system for a biotech diagnostic device at Portable Diagnostic Systems. Leads INIT Build teams from architecture to deployment.

Genuine Labs — Software Engineering Intern [May 2026 — Present]

• Designed and shipped a TypeScript eval runner that executes test suites against a live AI chat service, grades results, and produces structured pass/fail reports.
• Built seeded Firebase emulator test environments, deterministic graders, and multi-turn conversation case support.
• Adapted the pipeline across model providers via OpenRouter and ran the team's first eval baselines.

Portable Diagnostic Systems — Embedded Software Engineering Intern [Summer 2026]

• Designing the over-the-air update system for a portable diagnostic device running embedded Linux at a seed-stage biotech startup.
• Authored the OTA proposal and phased implementation plan: signed update bundles, A/B partitioning with automatic rollback, and durable audit logging.
• Engineered for intermittent connectivity, persistent state decoupling, and fail-safe operation in the field.

INIT Build — Team Lead [Spring 2026]

• Leading development of NoteBud, an AI-powered RAG study companion (LlamaIndex, Gemini, Next.js, FastAPI, PostgreSQL + pgvector, Docker, Google Cloud).
• Designed the full-stack architecture managing users, classes, notebooks, and source files end-to-end.

INIT Build — Team Lead [Fall 2025]

• Architected an end-to-end ML pipeline with CatBoost, LightGBM, and Temporal Fusion Transformer models.
• Owned all core backend services in Python: modular design, reproducible training runs, and rigorous cross-validation with metric tracking for time-series workloads.

AI4ALL — Fellow

• Built production-style ML models including Random Forest and XGBoost to forecast real-world signals with high predictive accuracy.
• Engineered features, performed hyperparameter tuning, and documented pipelines and results for future iteration.