Isaiah Murray

Resume

Isaiah Murray · Sophomore, Class of 2029 · U.S. Citizen · Embedded systems, instrumentation, and perception for physical hardware

Resume (PDF)

Experience

  1. 2026.03–2026.08Somerville, MA

    Formlabs

    Electrical Engineering Intern, SLS & Materials R&D

    • Built end-to-end powder-bed defect detection pipeline — GPU-accelerated classical CV region proposal feeding a ResNet-18 classifier over a nine-class taxonomy — replacing a binary detector with classified, severity-ranked results; surfaced order-of-magnitude defect-rate differences between machines running identical material.
    • Designed fleet-wide heater fault detection from first-principles physics after establishing no labeled fault data existed; root-caused a 9% false-positive rate to PWM phase jitter between hardware channels and restructured the check to 0% on healthy fleet data — caught a wiring fault on a unit 12h from shipping to a beta customer.
    • Replaced a manual filter-swap characterization procedure with a controlled-evaporation sweep producing a continuous pressure–flow surface in one run; built the resulting virtual flow sensor and validated it closed-loop against an independent sensor outside the control loop.
    • Built in-situ dielectric cure monitoring rig (PyQt6, VISA impedance analyzer, SSH-triggered exposure); identified cross-machine timing as the data-quality limit and added a dedicated fire-time recorder, enabling ~40 Hz transient capture across 300+ runs. Correlated ionic viscosity with FTIR, enabling FTIR emulation at ~300 Hz.
    • Built browser-native lifetime test runner (acoustic cycle detection, crash-safe resumable state, automatic Word/PDF report generation) and ran five qualification campaigns — longest logged 515 insertion cycles.
  2. 2024–2025Cambridge, MA

    Cherish Health

    Engineering Team

    • Automated phased-array radar calibration; replaced manual gain/phase tuning with an FPGA-controlled workflow using spectrum analysis and algorithmic optimization, cutting calibration from hours to minutes and reducing side lobe interference 20% (TX) / 30% (RX).
    • Built internal tooling for storing and comparing radar characterization data; accelerated state annotation throughput.
    • Sourced alternative LED supplier and coordinated contractors for molded fiber packaging, reducing BOM cost by 5%.
    • CAM'ed and manufactured injection-molded and machined components.
  3. 2023.05–2023.08Cambridge, MA

    MIT Chem-E — Strano Group

    Laboratory Assistant

    • Selected for MIT's HIP-SAT summer research program (Strano Lab); investigated graphene nanostructure additives in octadecane-based phase change materials to improve thermal conductivity and raise thermal resonator output voltage.
    • Presented findings at HIP-SAT Symposium and Boston Mammalian Synthetic Biology Symposium.
  4. 2022–2023Cambridge, MA

    Cherish Health

    ML Intern

    • Trained detection models and built data collection tooling in Python/TensorFlow.
    • Ran technical demos for CEO and CTO supporting a successful Series A round.

Additional experience

  1. 2024–2025Dorchester, MA

    Access Sport America

    Coach

    Adaptive watersports programming for athletes with disabilities; specialized in nonverbal autism support.

  2. 2016–2022Cohasset, MA

    Center for Student Coastal Research

    Student Researcher

    Won Marjot grant to study Zostera marina; built automated environment chamber testing recovery in Labyrinthula zosterae-infected specimens.

Projects

ProjectWhat it is
Autonomous Swarm in progress Multi-vehicle autonomy stack (Isaac Sim, Pegasus, PX4 SITL, MAVSDK-Python); four-vehicle concurrent autonomous flight in sim, hardware build underway on an STM32H743 FC.
Olin Baja Racing Ride-height suspension sensor for damper tuning, CAN-based data acquisition & diagnostic canopy.
CLASP BLE proximity-logging wearable on Nordic nRF52: custom PCB across multiple revisions, low-power RF design.
Bragi Real-time Quest 3 XR perception pipeline: YOLOv5 detection driving contextual overlays, Google Cloud STT/TTS.
Binaric Acoustic data transmission protocol: layered modulation, Manchester-clocked multi-tone encoding, adaptive error correction.

Education

  1. 2025–2029Needham, MA

    Franklin W. Olin College of Engineering

    B.S. Engineering, Electrical & Computer Engineering concentration · Expected May 2029 · GPA 3.75

    Built a facial recognition program, compared epidemiology models, and applied biomimicry to weight-efficient jumping mechanisms through project-based coursework.

  2. 2020–2024Cambridge, MA

    NuVu Innovation School

    8th–12th grades

    Project-based studios; Harvard Extension coursework in multivariable calculus, differential equations, linear algebra, CS50.

Skills

Languages
Python, C, C++, MATLAB, SQL, JavaScript
ML / CV
PyTorch, OpenCV, Kornia, scikit-learn, NumPy, Pandas, SciPy, TensorFlow
Hardware & Instrumentation
KiCad, Altium, PyVISA, USB DAQ, thermal imaging, nRF52, STM32, CAN, PX4/MAVLink
Tools & Infrastructure
BigQuery, Grafana, FastAPI, SQLite, Git, SolidWorks, Fusion 360, Blender, INAV, Docker, Linux

Interests

  • Rugby and MMA (Babson)
  • Babson Mazda Miata racing team
  • Classical guitar
  • Delta Tau Delta Alumni Chair