# James Morad

> Staff Machine eLearning Engineer at Block

Location: San Francisco Bay Area, United States
Profile: https://flows.cv/jamesmorad

Physics PhD using AI to help fight fraud on the Cash App.

## Work Experience
### Staff Machine Learning Engineer (Modeling) @ Block
Jan 2024 – Present | SF Bay Area

### Senior Machine Learning Engineer (Modeling) @ Block
Jan 2020 – Jan 2024 | SF Bay Area

### Data Scientist @ HeadSpin
Jan 2019 – Jan 2020 | Palo Alto
- Developed a reference-free video quality score model for predicting a user-perceived quality score time series for video using convolutional neural networks and decision tree-based models.
- Designed and created a video annotation framework that includes a Python backend for managing annotation experiments that interfaces with a cross-platform mobile app developed in React Native, enabling scalable aggregation of ground truth labels for video quality.
- Developed novel interpretable models for evaluating video quality.
- Maintained ownership over developed code with responsibilities that include unit-testing, production deployment, and code and model maintenance.

### Insight Data Science Fellow @ Insight Data Science
Jan 2018 – Jan 2018 | San Francisco
• Created a web app that presents users with relevant Instagram hashtags based on a user
submitted tattoo image.
• Scraped ~100,000 images and developed a pipeline for image classification using
convolutional neural networks and nearest neighbor clustering algorithms.
• Deployed application on AWS using nginx, gunicorn, and Flask.

### Senior Software Developer @ Cosylab
Jan 2017 – Jan 2018 | Palo Alto
• Built a laser steering control system using EPICS with actuators, energy sensors, and
cameras, enabling an O($10mil) experimental x-ray laser facility.
• Reviewed and delivered production code using git and cvs for versioning on development
systems.
• Heavy Python development of tools for aggregation and storage of data and in-house
Python and bash tools for automated software execution control.

### Software Developer @ Cosylab
Jan 2017 – Jan 2017 | Menlo Park
Worked on designing and developing  accelerator system controls software for SLAC National Lab.

### Graduate Student Researcher @ UC Davis
Jan 2011 – Jan 2017
• Introduced crowd-sourcing to the dark matter community by developing an in-house
Mechanical Turk-style web app for gathering labeled data to train a pulse classifier.
• Gathered data for and triaged problems with the LUX dark matter detector alongside a
team of ~100 international collaborators.
• Developed a full data acquisition and analysis suite providing new students a quick
understanding an experimental detector setup, reducing onboarding/productivity time
from months to days.
• Constructed feature selection algorithms for use in time series analyses allowing for signal
identification and classification.
• Developed software for data acquisition using embedded C, using direct memory access
for waveform retrieval from FPGA memory registers.
• Produced and deployed an in-house dark matter detector with challenges that included
data analysis pipelining, cryogenic hardware design, and electronic signal readout.

### Graduate Student Researcher @ Berkeley Lab
Jan 2015 – Jan 2016
• Constructed and deployed a small scale dual-phase xenon time projection chamber for use
in particle physics experiments
• Developed software for control, readout, and automation of detector vital systems such as
pressure, temperature, and mass flow
• Developed software for data acquisition and analysis
• Designed and constructed a readout circuit for a charge sensitive amplifier


## Education
### Doctor of Philosophy (Ph.D.) in Physics with a Designated Emphasis in Nuclear Science
University of California, Davis

### Master's degree in Physics
University of California, Davis

### Bachelor's degree in Engineering Physics/Applied Physics
University of California, Davis


## Contact & Social
- LinkedIn: https://linkedin.com/in/james-morad

---
Source: https://flows.cv/jamesmorad
JSON Resume: https://flows.cv/jamesmorad/resume.json
Last updated: 2026-03-29