22

Hi, I'm Zaid

Software Engineer, Entrepreneur, Machine Learning Engineer, Researcher

Zaid Nabulsi

Hi! I am a computer scientist and machine learning engineer. I'm currently a computer science student at Stanford University, interested in aritifical intelligence, machine learning, and computer systems. Please check out my portfolio, and feel free to contact me! I am always open to new opportunities.

Computer science is no more about computers than astronomy is about telescopes.

- Edsger Dijkstra

Education & Coursework

Stanford University

Expected to graduate in June 2020 with a BS and MS in Computer Science, with specializations in aritifial intelligence and computer systems. Current GPA: 4.03

Artificial Intelligence & Machine Learning
CS 221 Artificial Intelligence: Principles & Techniques
Dive into the world of aritifical intelligence, including topic such as Markov decision processes and constraint satisfaction.
CS 229 Machine Learning
Dive into the world of machine learning, covering a wide range of topics including SVMs, GLMs, ICA/PCA, and reinforcement learning.
CS 236 Deep Generative Models
dive into the world of generative models; including the probabilistic foundations and learning algorithms for deep generative models, including Variational Autoencoders (VAE), Generative Adversarial Networks (GAN), and flow models.
CS 124 From Languages to Information
A practical and applied course focused on extracting meaning, information, and structure from human language text, speech, web pages, social networks.
CS 274 Algorithms and Aritificial Intelligence in Computational Biology
Understanding and applying artificial intelligence to important sectors in computational biology.
CS 230 Deep Learning
A practical course useful for industry, focusing on the details of numerous different types of deep neural networks, their architectures, and their applications
CS 231N Convolutional Neural Networks for Visual Recognition
The details of neural net architecture, with a heavy focus on image classification and convolutional neural networks.
CS 224N Natural Language Processing with Deep Learning
Deep learning and neural network architecture with a focus on natural language processing.
CS 224U Natural Language Understanding
A project-oriented course focused on the task of teaching machines to understand and make sense of human language.
CS 199/Med 199 Undergraduate Research
Independent research for qualified students to undertake investigations sponsored by faculty. I worked in an artifical intelligence lab in the Stanford Medical School.
Algorithms & Systems
CS 106 Series Programming Abstractions
Deep dive into abstraction and software engineering principles ofmodularity. Focus on data structures and algorithms in C++
CS 107 Computer Organization & Systems
Fundamental concepts of computer systems, covering low-level programming, working from C down to the microprocessor.
CS 193X Full Stack Applications
Implementing full-stack web applications, with a focus on both client-side and server-side systems.
CS 140 Operating Systems
Comprehensive course on developing operating systems. In this course, students develop there own operating system (Pintos)
CS 110 Principles of Computer Systems
Principles and practice of engineering of computer software and hardware systems. Covers techniques for controlling complexity and strong modularity using client-server design. Also covers multiprocessing, multithreading, networking, virutal memory, security and encryption, and MapReduce
CS 161 Design and Analysis of Algorithms
Comprehensive course covering algorithm design techniques, and in-depth coverage of many exisiting and efficient algorithms. Emphasis on worst/average case analysis and recurrences/asymptotics
CS 143 Compilers
This course covers how compilers work and how they are developed. In this course, students write their own compiler for an academic programming language called COOL.
Mathematics
MATH 51 Linear Algebra & Differential Calculus of Several Variables
Geometry and algebra of vectors, matrices and linear transformations, eigenvalues of symmetric matrices, vector-valued functions and functions of several variables, partial derivatives and gradients, derivative as a matrix, chain rule in several variables, critical points and Hessian, least-squares, optimizations, and Lagrange multipliers.
CS 109 Probability for Computer Scientists
Basics of probability and the applications of probability in computer science. Focus on building machine learning systems, and the use of probability to analyze algorithms.
MATH 104 Applied Matrix Theory
Applications of linear algebra in science, engineering, and machine learning. Covers orthogonality, projections, spectral theory for symmetric matrices, the singular value decomposition, the QR decomposition, least-squares, the condition number of a matrix, and algorithms for solving linear systems.
CS 103 Mathematical Foundations of Computing
Proof-based course that covers propositional predicate logic, induction, grammars, automata, Turing machines, and NP-completeness.
Entrepreneurial
BIOE 70Q Medical Device Innovation
The application of design-thinking to the creation of healthcare technologies. Focus on crafting unique solutions to shape healthcare innovation through the use of technology.
URBANST 131 Social Innovation and the Social Entrepreneur
Introduces perspectives and endeavors of entrepreneurs and thought leaders who address social needs in the U.S. and internationally through private, for-profit and nonprofit organizations or public institutions.
ENGR 280 From Play To Innovation
Project-based course that is focused on the innovtion process itself, and enhancing it through playfulness. Use design-thinking to promote innovation in the corporate world, through working with real-world partners on design projects with widespread application.

Computer science is no more about computers than astronomy is about telescopes.

- Edsger Dijkstra

Work Experience

Google Brain

Software Engineering Intern (Jul-Oct 2019)

Worked with the Medical Brain team to develop deep learning models to predict/diagnose abnormalities

Developed a deep learning model that diagnoses abnormalities in chest x-rays through NLP and CV

Incorporated a pateint's past mammograms in deep learning models to help the models diagnose/predict breast cancer

Technologies used: Python, TensorFlow, C++, Go

Facebook

Software Engineering Intern (May-Jul 2019)

Worked with the Ads Ranking ML team to improve the quality ofthe ads ranking system.

Developed a machine learning model validator that ensured that machine learning models used in production met a certain criteria.

Turned the validator into a launch blocker, enhancing the quality of models used in the ads ranking ecosystem.

Technologies used: Python, Caffe, PyTorch, C++

Google Brain

Software Engineering Intern (Jun-Sept 2018)

Worked with the Brain team to improve the quality of Tensorflow Extended (TFX), and to prep for the TensorFlow 2.0 release.

Developed a deep learning classifier for YouTube live chats, improving AUC-PR from 0.73 to 0.78.

Technologies used: Python, TensorFlow, Go, C++

Documentation | Documentation | Article

Accountability Counsel

Software Engineering Intern (Jan-Apr 2018)

Developed a new webpage to make it easy to access and file complaints for world banks.

Designed an easy-to-use interface for the public to access.

Technologies used: Python, Django

Google

Software Engineering Intern (Jun-Sept 2017)

Worked on the App Maker Team to implement a brand new feature: Popups

Created a framework for adding cutomizable animations, from scratch, and enhanced srver-side rendering of the application.

Project is currently in prodduction, used by hundreds of thousands of clients around the world

Technologies used: Java, RPCs, Protos, SQL

Documentation | Video | Video

Stanford Code for Change

Software Developer (Oct 2016-Mar 2017)

Worked with a team of Stanford students to develop CultureMesh, a non-profit

Developed a Facebook Free Basics website in order to better help connect refugees

Technologies used: Python, Flask

Facebook TechStart

TechStart Role Model (Sep 2016 - Mar 2017)

Accepted to Facebook program to volunteer at public high schools to teach computer science

Organized lesson plans and events with teachers

Served as a mentor at Facebook's CS Education Week event at Facebook HQ in Menlo Park

Royal Industrial Plant

Electrical Engineering Intern (Jun-Sept 2016)

Collaborated with a team of 11 engineers to maintain the quality of 2 plants

Trained on the control and electrical requirements for GUI/Communication interface

Designed computer controls and instrumentation of machines

Computer science is no more about computers than astronomy is about telescopes.

- Edsger Dijkstra

Research


AI-Assisted Health CareSept 2018 - Present

Conducted research with the Partnership in AI-Assisted Care, which is an interdisciplinary collaboration between the School of Medicine and the Computer Science department focusing on cutting edge computer vision and machine learning technologies to solve some of healthcare's most important problems. I will be working under the guidance and mentorship of Fei-Fei Li, and my work will be focusing in on action recognition and computer vision, in health care settings.



Deep Learning ResearcherFeb 2018 - Present

Conducted research with the Laboratory of Quantitative Imaging in the Stanford School of Medicine. The lab uses computational methods for image classification and segmentation, with a focus on radioloy images. Worked on a cardiac disease classifier, based on heart radiology. Classifier first extracts image features, and then feeds them thorugh a CNN. Current work: Developing a model that predicts the occurrence of breast cancer, from mammography a few years prior.



Stanford School of MedicineMar 2017 - Feb 2018

Worked with a Stanford School of Medicine Labratory, with Dr. Ash Alizadeh and Dr. David Kurtz. The lab is focused on using non-invasive biologial biomarkers such as cell-free DNA (cfDNA), to detect cancer. I developed an innovative deep learning model that classifies a non-reference base as either a biological and possibly cancerous mutation, or as a human-introduced error, brought about by errors of the sequencer, or through other environmnetal exposures, such as oxidation of the base over time. This state-of-the-art model is currently undergoing comprehensive testing in clinical trials. Publication pending.




Disneyland will never be completed. It will continue to grow as long as there is imagination left in the world.

- Walt Disney

Projects


Lister Care Ltd

Developed a web and mobile application to help improve the experience of patients going into surgery. Application functionality includes an OR management system. The surgeon's team uses the web platform to communicate with the patient, who uses the mobile applciation. Automated messages are sent to the patient as reminders for the patient's upcoming operation. Beta version of the platform underwent clinical testing in Leicester Hospital in the U.K., showing an increase in patient satisfaction from 40% to 80%. Next step is to get the platform adoped by different departments. Technologies used: ReactJS, React Native, MongoDB, FireBase.



MRNGAN: Reconstructing 3D MRIS Using a Recurrent Generative Model

Part of a computer vision/machine learning research project for CS231N(Convolutional Neural Networks for Visual Recognition) with Professor Fei-Fei Li. Project conducted with a team of two other Stanford students. Paper is here.



RECURRENT PHRASE VECTORS: UTILIZING RECURRENCE IN WORD EMBEDDINGS

Part of a natural language processing/machine learning research project for CS224U(Natural Language Understanding) with Professor Bill MacCartney. Project conducted with a team of two other Stanford students. Paper is here.



A DL SOLUTION FOR BLOOD DIAGNOSTICS OF CANCERS THROUGH ERROR SUPPRESSION

Part of a deep learning/genomics research project for CS230(Deep Learning) with Professor Andrew Ng. Project conducted with a team of two other Stanford students. Paper is here.



Chatbot

Built a chatbot that recommends movies for each user, by first asking users to list 5 movies and their attitudes towards each. The backend is based on a recommendation system (linear algebra), with an exisitng database of users and movies. The chatbot is fully conversational. Chatbot developed in python.



VoiceMe

Developed a mobile application for iOS that allows users to record and manipulate audio to add effects to voice. Used Xcode and Swift to develop application.



Stanford Maps

Developed a web application for Stanford students to efficiently navigate hrough campus. Backend of the application calculates the most efficient route between locations on campus. Used the OpenMapStreet API to visually display results to the user on the front-end.

Disneyland will never be completed. It will continue to grow as long as there is imagination left in the world.

- Walt Disney

Contact Info

Mail: znabulsi [at] cs [dot] stanford [dot] edu Mail: zaidnab [at] gmail [dot] com Phone: (925) - 360 - 8625