About ME

no preview

Data Scientist at Roads & Transport Authority

  • Dubai, United Arab Emirates
  • +971 561788695
  • izk.abrhm@gmail.com

Accomplished Data Scientist with a passion for delivering valuable insights through analytical functions and data retrieval methods. Committed to helping companies advance by helping them to develop strategic plans based on predictive modeling and findings.

Achievements include:

  • Created a Geo Spacial AI Application which recommends new bus routes for public transport users within Dubai increasing overall revenue by ~ 400K AED
  • Implementing data regression models to project the number of days a transport bus can operate before the next maintenance.
  • Developed dashboards and tools to help visualize insights and patterns

    • Linguistic Ability

    English

    Fluent

    Malayalam

    Native

    Hindi

    Intermediate

    Interests

    Coding

    Travel

    Fitness

    Documenteries & Films

    Resume

    Education Credentials

    Master of Science in Data Science

    2019 - 2020

    Heriot-Watt University, Dubai

    Completed: Distinction Award

    • The Master of Science in Data Science program provided a comprehensive grounding in key areas like Statistics, Machine Learning, Big Data, Visualization, NLP etc.
    • Awarded UAE Golden Visa (Distinction Student Category) in recognition of academic excellence.
    • Completed dissertation by implementing an Image Description model.

    Programme Content:

    Masters Project & Dissertation: Image Description Model using multimodal space embedding
    Statistical Modelling & Analysis
    Data Mining & Machine Learning
    Biologically Inspired Computation
    Industrial Programming
    Applied Text Analytics
    Big Data Management
    Data Visualisation & Analytics
    Research Methods & Project Planning

    Bachelor of Technology in Information Technology

    2012 - 2016

    Rajagiri School Of Engineering & Technology, Kochi, India

    Mahatma Gandhi University, India

    Completed: First Class

    Certifications

    Dataiku: ML Practitioner

    2024 July

    Certified by Dataiku Learning Platform for mastery of ML tools

    Dataiku Advanced Designer

    2024 August

    Certified by Dataiku Learning Platform for mastery of advanced Data pipeline tools and concepts<

    GreyAtom

    2020 October

    5 Day Data Science Workshop by GreyAtom

    Data Science A-Z

    2020 December

    Real Life Data Science Exercised Included

    Professional Experience

    Data Scientist

    March 2021 - Present

    Roads and Transport Authority, Dubai, United Arab Emirates

    • Created a geo-spacial AI app that generated new travel routes recommended for customers using the Shail app for Bus transportation, maximizing overall travel options increasing revenue by ~ 400K AED.
    • Engineered highly accurate supervised learning predictive regression models to anticipate major internal system failures in RTA buses, resulting in a remarkable 15% enhancement in preventive maintenance operations for the RTA Bus department.
    • Developed innovative unsupervised learning classification models to evaluate & categorize the overall condition of buses upgrading the bus inspection process and achieving a notable 13% improvement.
    • Devised complex, cross-functional analytics solutions, and statistical analysis dashboards (MicroStrategy/Power BI) to continuously monitor and optimize key performance indicators (KPIs) empowering informed decision-making across RTA data infrastructure.
    • Initiated the automation of daily, weekly, monthly, and quarterly reports for internal stakeholders, streamlining the audit process and ensuring the timely delivery of critical information using Talend.
    • Leveraged Apache Spark and Kafka to build scalable data pipelines, streaming massive datasets (>1TB) from diverse sources like AWS/Azure Cloud networks, IOT Devices, APIs, PDW, Datalakes, Relational Databases, etc., enabling seamless data accessibility.
    • Optimized operational efficiency by automating tedious and repetitive data consolidation and extraction tasks, resulting in an impressive total weekly time savings of approximately 450 minutes.

    Data Science Intern

    January 2021 - Present

    Beinex Consulting, Kochi, India

    • Incorporated Optical Character Recognition application using OpenCV to extract text from images and PDF files reducing manual search by a margin of 37%.
    • Translated benefits of machine learning technology for non-technical audiences, including cost/benefit analyses.
    • Collaborated with senior personnel to define and meet data modeling standards for deep dive projects.

    Senior Systems Engineer

    February 2017 - June 2019

    Infosys, Bangalore, India

    • Engineered a user-friendly and swift e-commerce website for a retail client which reduced overall operating costs by 43% and improved efficiency by 57%.
    • Optimized and responsive Angular web pages were launched that seamlessly interacted with back-end API with a page load speed of ~1.2 seconds.
    • Successfully implemented various scenarios that come across in the front-end such as routing, tabs, graphs, etc. which received >90% positive user feedback .
    • Expedited planning, implementation, and integration of applications with clients, internal staff, and other stakeholders within 2 week sprints following Agile framework.
    • Moderated and analyzed code with Technical Architects to optimize performance and minimize faults in applications within 48 hours of each sprint.

    My Works

    Image Description model (CNN + LSTM)

    The Image Description model in this project consists of a CNN and LSTM network where the outputs from each neural network are aligned in a multimodal space embedding to generate sentence descriptions from a given image.


    Developed using: Python, TensorFlow, Keras, Scikit-Learn, Matplotlib, NumPy, Pandas, Tweepy

    Employee Churn Prediction

    This project is done to understand and analyze the employee patterns in leaving the company and predict if further employees tend to leave as well using Logistic Regression and Random Forrest Classifier


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Matplotlib

    Customer Segmentation Analysis Model

    This model which I have implemented using K-Means Clustering and PCA analyzes the banking patterns of customers and divides them into various groups allowing the marketing team to target customers with specialized promotions.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Tensorflow, Keras, Matplotlib

    Sales Forecasting Model

    In order to predict daily sales of the company, a time series sales prediction model using fbprophet has been developed. Before implementing the prediction model, the mandatory procedures like data collection, data cleaning, exploratory data analysis and visualization have been carried out.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Matplotlib, fbprophet

    Chest Disease Classification model

    A deep learning CNN model has been implemented to inspect X Ray images of human chests and classify whether the chests are healthy or are infected with diseases such as Covid-19, Viral Pneumonia or Bacterial Pneumonia.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Tensorflow, Keras, Matplotlib

    Amazon Sentiment Analysis Model

    I have implemented a Sentiment Analysis model on Amazon Alexa reviews. Using CountVectorizer along with Naive Bayes Classifier or Logistic Regression, we can predict whether their customers are satisfied with the product or not.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Matplotlib

    Image Segmentation Model

    I have automated the process of detecting and localizing defects found in Steel manufacturing using a ResNet50 model. The defect images pass through a ResUnet model which will be trained to identify and highlight the defective area.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Tensorflow, Keras, Matplotlib

    Advertising Regression Analysis

    In this project, A Linear Regression model has been implemented to predict sales from marketing/advertising spends on different platforms such as TV, Radio and Newspaper.


    Developed using: Python, NumPy, Pandas, Scikit-Learn, Matplotlib

    Skills

    • All
    • Hard Skills
    • Soft Skills
    • Programming Languages
    • Technologies and Frameworks

    Data Science

    Machine Learning

    Artificial Intelligence

    Neural Networks

    Data Analysis

    Descriptive/Predictive Analysis

    Data Cleaning

    Data Preprocessing

    ETL

    Data Visualization

    A/B Testing

    Model Deployment

    Statistical Analysis & Modelling

    Big Data Management

    Data Mining

    Natural Language Processing

    Data Warehousing

    Object-Oriented Programming

    Coding

    Data Structures

    Agile Methodologies

    Problem Solving

    Web Development

    Testing

    Debugging

    UX/UI Designing

    Multimedia Communication

    Python

    R Programming

    C++

    C

    SQL

    Java

    C#

    MySQL

    Cypher

    HTML5

    CSS3

    JavaScript

    Unix Shell

    Angular(2+)

    Dataiku

    Spark

    Informatica

    Hive

    Jupyter

    Apache Hadooop

    Apache Kafka

    ArcGIS

    Azure

    Anaconda

    TensorFlow

    PowerBI

    MicroStrategy

    Tableau

    Oracle RDBMS

    Bootstrap 4

    Microsoft SQL Server Management Studio

    SQL Server Integration Services (SSIS)

    JIRA

    PyCharm IDE

    D3.js

    SVN

    Git

    GitHub

    Angular-CLI

    Neo4j

    Weka

    Microsoft Office

    Microsoft Sharepoint

    Data Driven

    Analytical Mindset

    Descriptive Reasoning

    Fast Learner

    Clutch Performer

    Results Oriented

    Effective Communicator

    Team Collaborator

    Critical Thinker

    Creative designer

    Positive Attitude

    Troubleshooter

    Contact

    My Address

    Dubai

    Social Profiles

    Email Me

    izk.abrhm@gmail.com

    Contact Me

    +971 561788695