JAAFAR

ALABOOSI

Data Scientist & Machine learning engineer

PROFILE

Last semester student but fully

experienced in working on

machine learning and deep

learning engineering

Have an experience of full

stack machine learning and

deep learning

Working with AWS cloud and

services, like Sagemaker,

SAM, lambda functions

Familiar with different

platforms, frameworks, and

programming languages, like

Tensorflow, Docker, Bitbucket,

Python, R, MySQL, Django,

and CUDA

Able to build endpoint and

APIs to serve for the machine

learning and deep learning

models

CONTACT

PHONE:

+601*********

EMAIL:

adnc0j@r.postjobfree.com

EDUCATION

IIUM

2017 - PRESENT

Programming, Data science, Machine learning, Math. Information Technology Engineering

2013 - 2015

Data structure, statistics

WORK EXPERIENCE

Full stack machine learning engineer

NexMind AI Group

2021–present

Full time machine learning engineer

Sets plans and designing for ML & DL models

Gathering data

Preprocessing and feature engineering

Dealing with unstructured data

Building model or usage of transfer learning

Applying automated training and real time inferences using Docker and AWS cloud

Deploy models using Sagemaker, SAM and lambda functions on AWS cloud.

Evaluate and enhance models

Programing Tutor

IIUM

2017–2017

Teaching programming for bridging students

C++ Workshops

ICDL full course

LANGUAGES

Arabic & English

SKILLS

SOFT SKILLS:

High analytic skills

Working with data

Communication

Teaching skills

Hard working for long hours

Very fast learner

PROGRAMMING LANGUAGES:

Python

R

MySQL

Oracle

Java

JavaScript

C++

SOFTWARE ENGINEERING SKILLS:

Collect requirements

Set up plans

Features analysis

Pair programming

Testing and evaluation

FRAMEWORKS:

Tensorflow

Sagemaker SDK

Django

Boto3

TECHNOLOGIES AND PLATFORMS:

IBM Watson

AWS cloud

Sagemaker

Docker

Git

Bitbucket

Lambda Function AWS

SAM functions

Storages and policies on AWS

PROJECTS

Financial transactions fraud detection

It is both a research and a development project, which achieved new significant results in fraud detection taking in consideration all the 26 papers published on this topic.

It tests many models on different datasets, and deals with complex data problems and solve them using advanced sampling techniques.

Provides a genuine new idea of 2 layers model, each layer with 2 models, using deep learning and auto-encoders

Achieved significant accuracy of 99.4%

Search Engine optimized text generator using GPT-2

The goal of this project is to generate paragraphs or articles that are search engine optimized, so when these paragraphs are included in web page, it will achieve high rank

The idea is genuine, market analysis and planning was done

The project provides real time fine-tuning and real time inferences

It uses many advanced technologies, libraries and frameworks, where I have used Docker, GPT-2 API, Tensorflow, Bitbucket, CUDA, SAM and Sagemaker on AWS cloud.

I made the project fully compatible with the latest instants of GPUs provided on the cloud

I developed the project fully, in terms of providing all its functionalities, which includes: text crawling, preprocessing, automated training, automated deployment, text generation, S3 file storing, and returning back a cleaned results.

The project was developed with high abstraction, where changes and improvements can be done later without effecting other parts of the project

Covid-19 Analysis

One of the biggest personal projects that I published online as an open source, the project follows the full standard data science/ machine learning cycle.

It used 7 different big datasets to answer crucial questions about COVID-19

Many graphs and charts were built through this project

It was built using R language only

The project contains 73 pages done during one month

For the advanced analysis, the project uses machine learning models to come up with relationships between features and output, which is the stage of inferences

The project includes medical model and required some medical domain knowledge to build SIR model, which in our case became an SIS model, which is the stage of prediction Clustering Malaysian universities by market needs

The project uses python on IBM Watson cloud

No ready dataset was found, so I built my own crawler to collect data about Malaysian universities and clean them

The data was used to cluster the universities according to the services around them, using FourSquare API and show the clustering on a live map.

TECHNICAL PROJECTS

Building a neural network from scratch along with its optimization algorithms

Building a convolutional neural network from scratch

Building an object detection model from scratch

Building Face recognition model

Building an RNN along with its LSTM gate from scratch

Building all previous models with Tensorflow

Text classification model using Tensorflow

Image classification model using Tensorflow

CERTIFICATES AND COURSES

IBM Data Science Specialization (10 courses)

Data Science: Foundations using R Specialization (5 courses)

Tensorflow by DeepLearning.ai

Deep learning by DeepLearning.ai

Learning methods by McMaster & San Diego University

Web development

Training of Trainers (TOT) by Ministry of education

ComptiA A+

ComptiA Network+

Cisco (CCNA)

Public General Speaking

Business Administration

TESOL

Mobile Maintenance

Contact this candidate