Summary
Overview
Work History
Education
Skills
Websites
Certification
Languages
Personal Projects
Technical Skills
Objectives
Research
Timeline
Generic

Ali Rahbarimanesh

Dordrecht

Summary

Experienced data scientist with expertise in machine learning, predictive analytics and data modeling. Hands-on experience in building scalable solutions and managing data-driven projects. Primarily focused on data science, with a strong interest in exploring and contributing to other areas such as data analytics and data engineering. Familiar with cloud platforms, ETL processes and data pipelines, with a commitment to continuous learning and delivering impactful solutions across multiple domains.

Overview

3
3
years of professional experience
1
1
Certification

Work History

Data Scientist

Ista Azaran Sazeh
04.2019 - 05.2022

As a Data Scientist at Ista Azaran Sazeh, I developed machine learning models and data-driven solutions to optimize project timelines, construction costs, and resource allocation. Collaborating closely with project managers, construction teams, and executives, I provided insights that enhanced decision-making and project efficiency. Additionally, I contributed to the design of ETL pipelines to support data integration and maintain data quality.

  • Developed and deployed machine learning models to predict project timelines, construction costs, and resource allocation, increasing decision-making efficiency for project management teams.
  • Analyzed large datasets, including project management data, material costs, and building specifications, to provide actionable insights for construction teams and decision-makers.
  • Delivered insights through Power BI by creating interactive dashboards for managers and stakeholders.
  • Designed and maintained ETL pipelines, ensuring data integrity, seamless integration, and model validation to support key data engineering functions.
  • Collaborated with project managers and construction teams to align data-driven forecasts with project goals and client requirements.

Key Achievements

  • Reduced schedule estimation errors (MAE: 7 days) by leveraging various algorithms including Random Forest, LightGBM, and Neural Networks, ultimately selecting the most effective for improving project planning accuracy.
  • Achieved 5% MAPE in predicting building completion costs through an iterative approach using Linear and Lasso Regression alongside XGBoost, leading to optimized project budgets.
  • Applied both Random Forest and SVM to classify building types with 92% accuracy, enhancing planning accuracy for construction projects by testing multiple methodologies.
  • Forecasted material costs with 8% MAPE, utilizing a range of techniques such as Random Forest, CatBoost, and SVR to ensure optimal procurement and resource allocation.
  • Improved data processing efficiency by optimizing ETL pipeline components, streamlining processes, and boosting data quality.
  • Developed interactive Power BI dashboards that synthesized complex data into actionable insights, enabling executives to make informed project decisions.
  • Increased project profitability by 10% by employing data-driven insights and leveraging advanced models to guide resource allocation strategies.

Education

Master of Science - Data Science and Society (Business Track)

Tilburg University
Tilburg, Netherlands
01.2024

Pre-Master - Data Science and Society

Tilburg University
Tilburg, Netherlands
01.2023

Skills

  • Problem-solving
  • Cross-functional collaboration
  • Time management
  • Attention to detail
  • Proactive improvement

Certification

Data Science - Tose’e Institute

Languages

English
Proficient
C2
Dutch
Beginner
A1

Personal Projects

I have worked with various datasets in areas such as retail, healthcare, financial services and more, while my focus was real estate and construction. Below are some representative examples.

  • Transaction Fraud Detection and Analysis: Developed a fraud detection model using Random Forest and Logistic Regression, achieving 90% accuracy. Reduced false positives by 20% through feature engineering and visualized fraud patterns using Power BI, cutting response times by 15%. Simulated an ETL process for data pipeline management using Databricks.
  • Customer Behavior and Transaction Segmentation: Applied K-means and hierarchical clustering on an open-source dataset to identify customer segments, leading to a 10% increase in engagement. Visualized results with Power BI and processed data using Azure in an experimental setup.
  • Sales Data Analysis and Forecasting: Improved sales forecasting accuracy by 20%, reducing stockouts by 15%. Created Power BI dashboards contributing to a 10% increase in sales and a 5% efficiency boost. Simulated an ETL pipeline for data extraction and reporting.

Technical Skills

  • Programming Languages: Python, SQL, R
  • Machine Learning: Predictive modeling, clustering, forecasting
  • Data Visualization: Power BI, Tableau (working knowledge)
  • Cloud Platforms: Azure, GCP, Databricks (working knowledge)
  • Data Engineering: ETL processes, data pipelines, validation
  • Big Data Tools: Scikit-learn, Pandas, NumPy
  • DevOps: Apache Spark, Airflow, CI/CD, Azure Data Factory (working knowledge)

Objectives

I am looking to use my data science expertise in machine learning, predictive modeling and analytics to create impactful business solutions. I am eager to further explore all aspects of data, including data engineering and analytics, and strive to contribute to comprehensive data-driven strategies while continually learning new tools and technologies.

Research

  • Master's degree thesis: 'A Comparative Analysis of Machine Learning Algorithms for Predicting Time on Market (TOM) in Real Estate for Funda.nl'
    - Focus on regression models to predict time on market (TOM), an important factor in real estate transactions. A comparative analysis of machine learning algorithms (Random Forest, CatBoost, XGBoost, LightGBM) and conventional statistical models was performed using a novel feature selection method with RFECV and hyperparameter adjustment. This study showed which algorithms performed better in terms of prediction accuracy and provided new insight into the factors affecting TOM.

Timeline

Data Scientist

Ista Azaran Sazeh
04.2019 - 05.2022

Pre-Master - Data Science and Society

Tilburg University

Data Science - Tose’e Institute

Master of Science - Data Science and Society (Business Track)

Tilburg University
Ali Rahbarimanesh