Basel Mostafa Email & Phone Number

Menlo Park, Ca, Us

Worked on content moderation infrastructure for commerce content (e.g. product postings, ads) on Meta's surfaces.Projects include:1) Building a data pipeline for triggering product reviews based on user interaction with dynamic ads.2) Optimizing the team's review-triggering framework to detect triggering conditions earlier (keeping reviews in sync with user activity) and reducing load on the most pressing system bottleneck by 3x.3) Improving system visibility by logging more pertinent data and implementing dashboard maintenance pipelines.

Seattle, Wa, Us

Added configurable threat-detection capabilities on HTTP headers and cookies to the AWS Web Application Firewall (AWS WAF) product.Work includes:1) Implementing Nginx add-ons in C to run new threat-detection procedures for HTTP headers and cookies on requests before they are received by user applications.2) Robustly integrating the new features into the WAF infrastructure by updating the DB representations and binary encoding schemas of customer configurations to represent the new configuration options while ensuring backward and forward compatibility, updating checksums to track propagation of new configurations to edge nodes, implementing canaries for monitoring system health and triggering alerts.

Copenhagen, Dk

Investigated the implications of including temperature-dependence of the IMF (a galaxy's initial distribution of stars by mass) in models of galaxy formation. Presented at DAWN Summit 2020. Coauthored three papers published in The Astrophysical Journal.Built an efficient data pipeline for fitting IMF temperature to a catalog of galaxies, revising the estimates of various galaxy properties based on the new temperature fits, and analyzing and visualizing the effects on the star-forming main sequence (the empirical relationship between mass and star-formation rate of star-forming galaxies). Implemented optimizations to handle the large data volume and low system memory availability, including: 1) splitting the pipeline into a dependency graph of smaller modules whose execution is automatically scheduled by GNU make in order to facilitate persistent caching of expensive results and to reduce redundant computation, 2) leveraging pickle to maximize the data structure encoding efficiency.

Pasadena, California, Us

Worked on software for an autonomous submarine competing in RoboSub.Projects include:1) Tuning CRT's vision system to increase processing speed while maintaining precision. Achieved a 36% speed-up over CRT’s previous vision system configuration by enabling int-8 weight quantization, AVX, and SSE in the YOLO object detection model.2) Using OpenCV to track an underwater cart through a stream of images as seen by a moving submarine.

Led the development of the drive-train and game element manipulator used on FRC Team 2473's 2019 robot. Primary designer, scheduler, and planner for those projects. Ensured that my team had a clear understanding of expectations and held them to a high standard of quality and rate of work. Worked closely with elevator and software sub-teams to ensure compatibility. Of the 70+ students that pass through CHS Robotics' FRC division each year, was the only one in the last three years to have successfully led two large projects in a single build season, despite the classification of the latter as a "reach" project with a low probability of success.

Collaborated with teammates to develop ideas and plans to achieve team objectives. Designed in Solidworks and built an offseason ball shooter, animatronic, and robot chassis. Contributed to the design of a climber during 2018 build season.

In collaboration with two other high school students, mentored a team of ten middle school students through two FIRST Lego League seasons. Coordinated discussions to assess team goals. Created and enforced schedules to build/code a robot, complete a research project, and build strong team unity in preparation for competition. The teams were successful and had a positive and memorable experience.

Led all efforts relating to the creation of FTC Team 11467's robot for the Velocity Vortex challenge. Delegated responsibilities to three mechanical/electrical teams and two software sub-teams, and regularly checked in with sub-team leads to ensure adequate progress and to make plans for correction as necessary. Designed all hardware prototypes and robot components in Solidworks with input from team and mentors. Developed drive code to vary sensitivity across the joystick in an ergonomic manner. Ensured that every team member had a meaningful role and expanded the scope of the team with new projects to suit team members' strengths and interests. At the Cisco Qualifier, advanced to semi-finals and won the Inspire Award, the highest honor awarded at the event.

Bachelor Of Science - Bs, Computer Science

High School Diploma