NM / Issue 01
Menu

About / engineering profile

Nathaniel Mapaye

Nathaniel Mapaye is a systems and embedded engineer based in Santa Cruz, California. He develops SysLib, a C++20 systems library with an SPSC ring buffer and a blocking MPMC queue, and EmbNode, a C++17 telemetry prototype with FreeRTOS task scaffolding and host-tested packet, power-accounting, and OTA coordination code. He is pursuing a B.S. in Technology & Information Management at the University of California, Santa Cruz, with graduation expected in June 2027.

Full name: Nathaniel Fransiscus Mapaye. My work also includes application security, Kubernetes, and AI evaluation.

nmapaye@ucsc.edu · GitHub · LinkedIn

Technical focus

Selected projects

EmbNode

A C++17 embedded telemetry prototype with FreeRTOS task scaffolding for ESP32 and STM32, host-tested packet framing, power accounting, and OTA coordination.

Problem
Intermittent networks and low-power constraints make telemetry delivery unreliable.
System design
FreeRTOS task scaffolding separates simulated sampling, aggregation, and transport hooks.
Evidence
Host contract tests exercise packet validation, power accounting, and OTA backend failures.
Outcome
A testable telemetry foundation; hardware DMA, network delivery, and field reliability still need validation.

View EmbNode sourceRead the telemetry design

SysLib

A header-only systems library with a lock-free SPSC ring, a blocking MPMC queue, and RAII resource wrappers.

Problem
Concurrent data-structure experiments need a small surface that makes correctness work explicit.
System design
Header-only C++20 SPSC atomics and a mutex-based MPMC queue keep ownership explicit.
Evidence
Tests cover wraparound, move-only values, wakeups, value lifetime, and concurrent delivery.
Outcome
A focused library for examining queue behavior, with sanitizer build options and explicit limitations.

View SysLib sourceRead the concurrency design

GitOps

A Go Kubernetes operator that turns service-level signals into automated canary decisions.

Problem
Canary releases need a clear decision path when service health changes.
System design
A Go Kubernetes operator combines Prometheus SLO signals with Gatekeeper policy checks.
Evidence
The rollback path triggers in under 30 seconds on a p95 regression.
Outcome
Release decisions become repeatable rather than dependent on manual monitoring.

View GitOps source

AURORA

A React Native iOS app unifying caffeine logging, intake dashboards, and psychomotor-vigilance testing through optional read-only Apple Health access and local-first storage.

Problem
Personal health signals need useful feedback without exporting sensitive data by default.
System design
React Native and Swift connect a local data model, optional read-only HealthKit access, and SQLite persistence.
Evidence
Caffeine logging, intake dashboards, and a 60-second vigilance test share normalized local inputs.
Outcome
A private iOS and iPad experience turns low-level data choices into a clear daily workflow.

Read AURORA case studyView AURORA demo

Experience

  1. LLM Specialist, Frontier AI Lab

    Jun 2026 - Aug 2026 · San Francisco, CA / Contract

    • Evaluated 520+ frontier AI outputs across software engineering and technical reasoning against correctness, instruction following, and edge-case criteria, producing reproducible evidence for model-improvement workflows.
    • Diagnosed 70+ recurring failure patterns and wrote evidence-backed annotations; 89% of reviewed cases could be rerun with the same behavior.

  2. Cyber Security AI Red-Teamer Intern, CFX Indonesia

    Jun 2026 - Present · Jakarta, Indonesia

    • Assessed 30+ assets across web, network, and control workflows in three workstreams: risk analysis, control documentation, and vulnerability remediation.
    • Analyzed and triaged 27+ vulnerabilities in OJK-regulated digital-asset infrastructure, tracking each finding through remediation with a clear evidence trail.

  3. Subject Matter Expert (SME), AI Alignment, Handshake

    Apr 2026 — Present · Remote / Contract

    • Audited 8+ repository-level software tasks across code, diffs, tests, and requirements to determine whether proposed changes satisfied task requirements.
    • Documented reproducible failure modes and implementation gaps under structured rubrics for AI-alignment review.

  4. Linear Algebra Tutor, UCSC ACE

    Jan 2026 — Apr 2026 · Santa Cruz, CA / Hybrid

    • Tutored students in linear algebra through guided problem-solving — systems of equations, matrix operations, vector spaces, linear transformations, eigenvalues, and diagonalization.
    • Led collaborative review and practice sessions focused on exam prep, homework strategy, and conceptual gaps; helped students move from memorized procedures toward step-by-step reasoning.
    • Adapted explanations to different skill levels and encouraged consistent study habits in a seasonal academic role.

  5. Data Center Engineering Intern, Bitera D.C

    Aug 2025 — Sep 2025 · Jakarta, Indonesia

    • Verified 16+ high-severity web findings through authenticated testing and retesting, producing actionable remediation evidence.
    • Supported five live data-center domains across Linux, networks, power, cooling, and alarms while documenting repeatable operational checks.

  6. Cyber Security Analyst Intern, Xapiens Teknologi Indonesia

    Jul 2024 — Sep 2024 · Tangerang, Indonesia

    • Executed authenticated scans and manual web tests on 50+ endpoints, validated exploitability, and produced proof-of-concept reports for several CVSS 8.0+ findings.
    • Prioritized remediation and documented retest evidence so client teams could track high-risk findings through closure.

Education

B.S., Technology & Information Management (TIM), University of California, Santa Cruz, Expected Jun 2027