HELLO, I'M
Ali Sanaei.
Data Scientist at KPN
Ali Sanaei
Data Scientist / AI Engineer
​
Rotterdam, Netherlands
alisanaei.n AT gmail.com
About
MY BACKGROUND
Drawing from 8 years of experience as a data scientist in multiple industries (telecommunications, banking, consulting and healthcare), I make data work for the problem at hand. Prior to that, I worked for 3 years as a product owner/project planner, working with Agile methods based on PRINCE2 and ITIL methodologies.
Over the years, I have developed expertise with multiple tools, including Python (Pandas, Numpy, Scikit-learn, PyTorch), Oracle PL/SQL, Microsoft SQL Server, Dataiku DSS, AWS Cloud (Certified AWS Cloud Practitioner), Microsoft Azure ML (Certified Azure Data Scientist Associate (DP-100)), R, Tableau, Power BI, etc. I have good communication and presentation skills, and the ability to explain technical concepts and results to a non-technical audience and stakeholders.
I am a forward-thinking individual, always eager to learn and adapt to new technologies and methodologies. As such, I am excited about the opportunities that the ever-evolving data science field presents, and I am committed to leveraging my skills to drive positive change.
​
Interests:
-
Data Science, Data Analytics, Data Visualization, Big Data, Machine Learning
Education
WHAT I’VE LEARNED
2020–2022
Eindhoven University of Technology (TU/e) – Eindhoven, Netherlands; Class of 2020
Master of Science; Data Science in Engineering
​​
GPA: 7.5/10
​​
Thesis Title: Evaluation of Unsupervised Anomaly Detection (EUAD) Algorithms In Collaboration with ING Supervisor: Prof. M Pechenizkiy
​​​
Thesis Abstract:
In several applications, such as fraud detection, it is crucial to identify anomalies in data. Different evaluation metrics, such as precision, recall, and F1 score, can be used to evaluate the performance of a machine learning algorithm. However, using these metrics is possible when sufficient labeled data is available. This research aims to evaluate unsupervised anomaly detection (EUAD) algorithms when no label is available.
Based on what is done in the literature, we see more potential in using
Mass-Volume (MV) curve and Excess Mass (EM). So, this study’s primary
research question is, “Are EM and MV capable of evaluating the performance of unsupervised anomaly detection (EUAD) algorithms?”. The answer to this question is far from straightforward. Hence, three different sub-questions are described to assess the quality of these evaluation metrics.
An evaluation framework is designed to evaluate the performance of EM and MV. The framework consists of three main parts: Datasets, algorithms, and metrics. In total, nine synthetic datasets and four real-world datasets are used. Regarding the algorithm, four popular anomaly detection algorithms (Isolation Forest, One Class SVM, kNN, LOF) with their different sets of hyperparameters comprising 213 algorithms are used. Alongside EM and MV, several ground truth metrics, such as precision, recall, and F5, are used. Besides, four hypotheses are considered for each evaluation metric to address the sub-questions.
In this study, we argue that it is unjustified to state that EM and MV are
capable of evaluating unsupervised anomaly detection algorithms by using several ground truths, a diverse range of anomaly detection algorithms, and a broad collection of datasets. However, our findings suggest EM can be useful in some domains or application scenarios.
​
Selected Courses: Advanced Process Mining, Advanced Algorithms, Foundation of Data Mining, Visualization, Data Modeling and Databases (using SQL), Longitudinal Data Analysis, Machine Learning (using Python), Big Data Management, Deep Learning (using Python), Algorithms and Data Structure, Logic and Set Theory
Experience
WHERE I’VE WORKED
Sep 2022–Present
KPN – Rotterdam, Netherlands
Mid-Senior Data Scientist / AI Engineer
​
In Corporate Insight & Analytics (CIA), I am part of the AI4Corporate team responsible for building data science solutions for stakeholders from Finance and HR.
Accomplishments:
• Managed the high-impact customer acceptance project: Developed machine learning models to predict fraudulent users through an early warning system incorporating responsible AI, resulting in a significant reduction of 2M euro revenue loss.​
• Developed an in-house B2B credit scorecard model: Delivered a machine learning solution to flag high-risk customers, reducing costs and increasing transparency in enabling proactive risk management and credit decisioning.
• Developing an agentic AI solution with integrated RAG functionality to automate and streamline the HR Requests for Advice (RFA) process, achieving a €120K cost reduction while improving efficiency and simplifying workflows.
• Implemented a Generative AI solution using Azure to automatically categorize and summarize employee survey feedback, cutting down manual work by 900 hours per year.​
• Contributed to the team's achievement of successfully reduce error in forecasting B2C transactions by 10% through the implementation of new time-series modeling.
Responsibilities:
• Led end-to-end AI / ML model development lifecycle, encompassing data collection, preprocessing, feature engineering, model building considering fairness and explainable AI, testing, and deploying to production.
• Engaged in data-driven product development by performing A/B testing to optimize key business metrics.
• Collaborated closely with stakeholders, understanding their business challenges and needs, translated them into feasible data problems, and provided data-driven solutions to support strategic decision-making.
• Collaborated effectively with cross-functional teams, including data scientists, data engineers, and machine learning engineers, to develop and execute data-driven strategies.
Tools:
• Proficient in Python, SQL, and Dataiku for data manipulation, analysis, and visualization.
• Hands-on in Prompt Engineering, Gen AI and other LLM models.
• Experienced in using Power BI for data visualization.
• Skilled in utilizing AWS and Azure cloud computing platforms.
Nov 2022–Aug 2022
ING – Amsterdam, Netherlands
Data Science Model Validator
I worked as a data scientist as part of the data science model validation team in the model risk management domain.
Achievements:
• Developed a framework to effectively evaluate the performance of unsupervised anomaly detection machine learning (ML) algorithms.
Responsibilities:
• Worked on the evaluation of unsupervised anomaly detection machine learning (ML) algorithms.
• Performed model validation on machine learning models in fraud detection and KYC domains.
Tools:
• Proficient in Python for data manipulation, and analysis.
• Experienced in using Tableau for data visualization.
Feb 2021–August 2021
Tradesparent – Rotterdam, Netherlands
Business Data Scientist Intern
Implemented an outlier detection module in SQL using statistical data analysis (Using Python for data preprocessing and Tableau for data visualization).
2016–2018
Tarbiat Modares University (TMU) – Tehran, Iran; Class of 2016
Master of Science; Industrial Engineering
GPA: 4.0/4.0 (18.23/20.00) Thesis Grade: 20/20 (Perfect)
Thesis Title: Designing a Big Data P4 Medicine based Registry (P4MR) Framework: Data Analytics and Visualization of Iranian Maternal and Neonatal’s (IMaN) Registry Supervisor: Prof. MM Sepehri
Thesis Abstract:
The objective of this study is to provide a framework for designing and implementing a Big Data registry, based on the international and domestic standards such a way that can be considered as the basis of developing prospective registries. This framework is designed based on the P4 Medicine concept, and it aims to incorporate the P4 Medicine features into the registry design and improving that. Another objective of this study is data analytics and visualization of the Maternal and Neonatal’s health registry to enhance the quality and accuracy of healthcare decisions and provide a data-driven decision support system.
The registry data for a birth certificate for five years (2013 to 2017) includes 7 million records and 220 fields. Data cleansing operation is done, then different data mining algorithms are applied on cleaned data for predicting diseases and that algorithms are evaluated. Afterward, data visualization was carried out according to the requirements of the neonatal department of the Iranian Ministry of Health, and its outputs are reported in the research.
The designed framework can be the basis for designing Big Data registry to consider registry design limitations and considerations, and registries are created concerning interoperability, integrity, and dimensions of data quality and privacy considerations, and prevent each registry from performing like different silos. Also, the use of data analytics and visualization can lead to create decision alternatives in the field of healthcare and reduce errors and improve quality and security in the mentioned area.
​
Selected Courses: Data Mining, Seminar, Engineering Statistics, Research Method, Simulation
​
​
Academic Projects (Graduate Program):
​
P-Hub Median Facility Location Problem, Industrial Systems Design Course (Dec 2016-Feb 2017)
Effect of measurement errors in the regressors, Engineering Statistics Course (Jan 2017-Feb 2017)
Clustering Techniques for Big Data Analysis, Data Mining and Knowledge Discovery Course (Mar 2017-Apr 2017)
Forecast the Onset of Diabetes Mellitus, Data Mining and Knowledge Discovery Course (Apr 2017-Jul 2017)
Global Logistics Development Strategies, Supply Chain Management Course (Apr 2007-May 2017)
The Simulation of Hasheminejad hospital Queuing System, Computer Simulation Course (Mar 2017-Jul 2017)
Application of queuing systems in production-inventory systems, Queuing Systems Course (May 2017- Jul 2017)
Big Data in Healthcare, Seminar Course (Apr 2017-Sep 2017)
Supplier Selection using MCDM Methods, Multiple Criteria Decision Making Course (Nov 2017-Dec 2017)
Dec 2017–Sep 2020
Hamrahe Aval (MCI) (Mobile Telecommunication Company of Iran) – Tehran, Iran
Data Scientist
I was working as a data scientist within the data hub of the sales & marketing department at MCI.
Accomplishments:
• Applied machine learning techniques to customer segmentation/classification that reduced the rate of customer churn by 5%.
Responsibilities:
• Conducted data analysis to make business recommendations (cost-benefit, invest-divest, forecasting, impact analysis) using different machine learning algorithms.
• Worked on data modeling, metric development, insights generation and recommendations using data mining and statistical techniques handling gigabyte and terabyte-sized datasets from internal sources such as CDR, CRM and external sources such as GSMA Intelligence.
• Delivered effective presentations of findings and recommendations to multiple levels of stakeholders, creating visual displays of quantitative information.
• Collaborated with cross-functional stakeholders to understand their business needs, formulate and complete end-to-end analysis that includes data gathering, analysis, ongoing scaled deliverables and presentations.
Tools:
• Proficient in SQL, R, RapidMiner, and IBM SPSS Modeler (Clementine) for data manipulation, analysis, and visualization.
• Experienced in using Tableau for data visualization.
About MCI:
Mobile Telecommunication Company of Iran (MCI) also known under its brand name Hamrahe Aval is the first and largest mobile operator in Iran. MCI is a subsidiary of the Telecommunication Company of Iran with approximately 19 million active postpaid and 49.4 million active prepaid subscribers. Hamrahe Aval's service is available in 1,239 cities and over 70,000 kilometers of highway in Iran. It provides roaming services via 270 partner operators in 112 countries. In December 2010, 5.5% of the MCI shares were offered on the Iranian over-the-counter market (Farabourse), at a value of $396 million which was the largest IPO-to-date in the Iranian OTC equity market. In August 2013, the company moved from the OTC to the Tehran Stock Exchange mainboard. The market value of the company in march 2017 was $ 4,326 million.
Dec 2016–Jan 2018
Gita Data Mining Research Institue – Tehran, Iran
Data Analyst
​
-
Performed data analysis and worked with the business to confirm, identify and document business logic for data mapping.
-
Extracted and analyzed data based on results using a variety of techniques, ranging from simple data aggregation via statistical analysis to complex data mining.
-
Produced ad hoc analysis and data modeling using IBM SPSS Modeler to understand and predict behavior trends to drive decision-making.
-
Communicated and disseminated the results of information with the business stakeholders.
​
Gita data mining company is the introducer of data mining in Iran known as the pioneer of data mining in Iran.
2011–2015
Iran University of Science and Technology (IUST) – Tehran, Iran; Class of 2011
Bachelor of Science; Industrial Engineering
GPA: 3.1/4.0 (15.13/20.00)
Final Project: Cloud Computing in Big Data
Supervisor: Prof. M Ghazanfari
​
Selected Courses: Computer Programming, Probability & Statistics, Calculus, Linear Algebra, Mathematics, Management Information Systems (MIS), Operation Research, System Analysis, Statistical Quality Control, Game Theory, Big Data, Project Planning & Management
​
​
According to the results of 2016 Times Higher Education ranking of Asian universities issued on June 20, Iran University of Science and Technology stands at 57th place among 200 top Asian Universities, promoting 12 stands since 2015. Five performance indicators as: teaching, research, citation, international outlook, and industry income are considered by THE for university rankings.
Jun 2013–Oct 2016
TOSAN (Banking and Payment Solutions Provider) – Tehran, Iran
Product Owner / Project Planner
​
As a Product Owner within APM (Agile Project Management) team, I am responsible for ensuring that our product meets the needs of our customers. This involves working on the product backlog, gathering feedback from users and stakeholders, and collaborating with my two Scrum teams to develop new features.
My responsibilities are:
-
Developed or updated project plans for information technology projects including information such as project objectives, information specifications, and schedules based on PRINCE2 and ITIL methodology.
-
Controlled project execution to ensure adherence to budget, schedule, and scope using Agile methods such as Scrum.
-
Developed work breakdown structure (WBS) of information technology projects.
-
Prepared project status reports by collecting, analyzing, and summarizing information and trends.
​
About Tosan:
TOSAN Corporation was founded in 1998 specializing in the provision of end-to-end banking solutions for both conventional and Islamic financial institutions. TOSAN is the first and largest market-leading provider of banking software solutions to retail, corporate, private, microfinance banks in Iran.
Operating in more than 17 years, we offer a diverse range of solutions encompassing all major components of the banking process. TOSAN group has served to over 37 financial institutions worldwide where is 9 in Asia and Pacific and 2 Banks in Africa, 1 in Europe and 1 in South America and 25 in the Middle East. TOSAN products have been deployed in the medium-size banks to large-scale credit institutes performing in over 5,000 branches and over 22,000 of tellers and over 50 million accounts along with over 25 millions of customers.
TOSAN group, which currently comprises of 14 companies, is active in a diverse range of banking systems such as software, hardware, banking solutions, payment services etc.
TOSAN recognized by Banking CIO Outlook magazine as An annual listing of 10 companies in 2016 that are at the forefront of providing retail banking solutions and impacting the marketplace in the world.
Digital Magazine Link: https://lnkd.in/bh4ziiZ
Extracurricular Activities
WHAT I’VE DONE
Sep 2021– Feb 2022
Eindhoven University of Technology – Eindhoven, Netherlands
Graduate Teaching Assistant​
​
"Data Mining and Machine Learning" & "Business Insight" Courses Learned to be more organized and responsible.
Jan 2021– Nov 2022
Mathematics & Computer Science Department Council, Eindhoven University of Technology – Eindhoven, Netherlands
Board Member
​
The task of the department council is to advise the department board and check the more significant decisions they make. I am the chair of the Communication Committee and a member of the Education and Well Being Committees
Learned to lead a team, be goal-oriented and define a roadmap.
Nov 2016– Oct 2018
Tarbiat Modares University – Tehran, Iran
Study Association Board Member​
​
Learned time-management and team working.
May 2012– Jun 2013
Iran University of Science and Technology – Tehran, Iran
Study Association Board Member​
​
Feeling as a part of a community and learned to express my ideas clearly, project scheduling and planning.
Publications
Quality Improvement through Intensive Care Registries: The Value of Big Data in Clinical Decision Making
Research
WHAT I'VE DONE
Journal Papers
2019
-
Sanaei Ali, Sepehri Mohammad Mehdi - Quality Improvement through Intensive Care Registries: The Value of Big Data in Clinical Decision Making, Dec 2019, Archives of Anesthesiology and Critical Care (AACC)
Conference Proceedings
2018
-
Sanaei Ali, Sepehri Mohammad Mehdi - Intensive Care Observational Registries: Big Data Analytics as a clinical decision support for critical care medicine, Oct 2018, International Congress on Critical Care Medicine: Systems, Technology & Management (ICCCMSTM), Tehran, Iran
-
Sanaei Ali, Fayaz Mohammad - How will GDPR affect the telecom business, May 2018, South Asian Telecommunications Regulators’ Council (SATRC), Tehran, Iran
2017
-
Sanaei Ali, Sepehri Mohammad Mehdi - Big Data in Municipal Healthcare Systems-Big Data and P4 Medicine, Mar 2017, Computer Society of Iran, Tehran, Iran
Skills & Test Scores
WHAT I BRING TO THE TABLE
Python
Machine Learning (ML)
Azure ML (Certified Data Scientist Associate)
Tableau
SQL
Dataiku Data Science Studio (DSS)
AWS
Power BI
R
IELTS Academic: C1 Level
GRE General - Quantitative: 166/170 (89th Percentile)
Awards
WHERE I SHINE
-
Ranked 3rd in "Hacking Mental Health" Hackathon, FruitPunch AI, Mar 2021
-
ALSP and Holland Scholarship for tuition fees and living expenses, Eindhoven University of Technology (TU/e), Sep 2020 - Aug 2022
-
Ranked 3rd GPA among the graduating class of 2016 in the Master Program, Tarbiat Modares University, Feb 2018
-
Government Tuition-fee scholarship for MSc degree, Sep 2016-Sep 2018
-
Ranked within the top 2% of participants in Iranian National University Entrance Exam for Master’s degree in Industrial Engineering, Sep 2016
-
Government Tuition-fee scholarship for BSc degree, Sep 2011-Sep 2015
-
Ranked within the top 1% of participants in Iranian National University Entrance Exam for Bachelor’s degree, Sep 2011
Interests
OUT OF OFFICE
Photography
Travel
Table Tennis
Tennis
Dance
Technology
Certificates & Courses
WHAT I'VE TAKEN
-
Dataiku Generative AI Practitioner, Dataiku
-
Stakeholder Engagement, SUAS
-
Microsoft Certified: Azure Data Scientist Associate, Microsoft
-
Managing Project Stakeholders, LinkedIn
-
Microsoft Azure Data Scientist Associate (DP-100) Specialization, Coursera
-
Build and Operate Machine Learning Solutions with Azure, Coursera
-
Dataiku MLOps Practitioner, Dataiku
-
Dataiku Advanced Designer, Dataiku
-
Dataiku ML Practitioner, Dataiku
-
Functions for Manipulating Data in SQL Server, DataCamp
-
Writing Functions and Stored Procedures in SQL Server, DataCamp
-
Intermediate SQL Server, DataCamp
-
Python Programming Course, Sematec
-
The Data Scientist’s Toolbox, Johns Hopkins University on Coursera
-
Big Data Modeling and Management Systems, University of California San Diego on Coursera
-
Introduction to Big Data, University of California San Diego on Coursera
-
Linux Fundamentals & Basic Administration (LPIC-1), Sematec
-
Data Mining and Business Intelligence Workshop, 3rd Conference on Industrial Engineering (CIE) - Sharif University of Technology and Amirkabir University of Technology
-
Internet of Things Workshop, IOT Academy (License Number: IoTA16-IN1224598)
-
Techniques and Concepts of Big Data, LinkedIn
-
Introduction to Data Analysis with Python, LinkedIn
-
Hadoop Fundamentals, LinkedIn
-
Introduction to Data Science, LinkedIn
-
Transitioning from Data Warehousing to Big Data, LinkedIn
-
Mobile networks overview (2G/3G/4G), MCI
-
Microsoft Excel Advanced, Tarbiat Modares University
-
Business Intelligence (BI) using Excel, Sharif University of Technology
-
PRINCE2 Foundation, Raydana
-
IT Infrastructure Library (ITIL® v3) Foundation in IT Service Management, Raydana
-
PMBOK (Project Management Body of Knowledge), Iran University of Science and Technology
-
Metaheuristic Algorithms in MATLAB, Tosee
-
Information Security Management System (ISMS), IMQ
-
Leadership & Team Management (Pacific Game), Gamelearn
-
Staff Empowerment development, Industrial Management Institute (IMI)