Nedim is available for hire
Hire NedimNedim Bayrakdar
Verified Expert in Engineering
Machine Learning Developer
Location
Amsterdam, Netherlands
Toptal Member Since
March 23, 2022
During his time at reputable companies such as ING, Heineken, and Adyen, Nedim has gained experience in both leading data science projects and in contributing to all stages of a data science project such as ideation, managing stakeholders, prioritization, evaluation of the projects, and creating and deploying models.
Portfolio
Consumer Facing Financial Assets Company
A/B Testing, Tableau, SQL, Amazon S3 (AWS S3), Machine Learning, Statistics...
Adyen
PySpark, Apache Airflow, LightGBM, Generalized Linear Model (GLM), Python...
Heineken
Python, Apache Airflow, Azure ML Studio, CI/CD Pipelines...
Experience
Availability
Part-time
Preferred Environment
Python, MacOS, Ubuntu, Visual Studio Code (VS Code), PyCharm
The most amazing...
...thing I've worked on is a model that optimized millions of payments per day in real time, leading to millions in additional yearly revenue for our clients.
Work Experience
Lead Data Scientist
2022 - PRESENT
Consumer Facing Financial Assets Company
- Developed a real-time dynamic pricing solution for financial assets within a consumer-facing application, combined with designing and running AB tests and experiments to iterate upon solutions leading to a 20% uplift in revenue (±10 million USD per year).
- Developed an A/B testing framework to design and implement A/B tests to make data-driven improvements to our consumer-facing product. Guided teams in running experiments.
- Built dashboards for tracking key experimental results and user metrics and behaviors, demonstrating value and insights organization-wide.
Technologies: A/B Testing, Tableau, SQL, Amazon S3 (AWS S3), Machine Learning, Statistics, Python, Artificial Intelligence (AI)
Lead Data Scientist
2021 - PRESENT
Adyen
- Led the optimization of online shopper conversion and payment authorization rates by using machine learning models that choose the optimal form of authentication per individual transaction in our online payment authentication product.
- Set and managed the priorities for the data science projects related to our online payment authentication product.
- Guided and mentored multiple junior and medior data scientists on our projects.
Technologies: PySpark, Apache Airflow, LightGBM, Generalized Linear Model (GLM), Python, A/B Testing, SQL, Big Data, Artificial Intelligence (AI)
Data Scientist
2020 - 2021
Heineken
- Developed a supply chain application that forecasted warehouse inventory for multiple products and provided recommendations on what and when to order via a Power BI dashboard, significantly reducing storage costs and wastage.
- Provided actionable insights to optimize the space utilization of shipment trucks, leading to savings of approximately one million euros per year.
- Automated revenue forecasting, resulting in a significantly reduced manual workload of the financial function.
Technologies: Python, Apache Airflow, Azure ML Studio, CI/CD Pipelines, Bayesian Inference & Modeling, K-means Clustering, Generalized Linear Model (GLM), LightGBM, Azure DevOps, SQL, Artificial Intelligence (AI)
Data Scientist
2018 - 2019
ING Group
- Implemented a real-time inference machine learning solution for detecting subscription and recurring payments for a mobile money management application (fintech).
- Contributed to algorithmic bond trading by using probabilistic models to determine prices.
- Researched analytical methods for approximating machine learning model uncertainties.
Technologies: Python, Data Science, Docker, Flask, PySpark, Big Data
Experience
Real-time Online Payment Authentication Optimization Model
Due to recently introduced regulations, online payments must be authenticated in the European Economic Area (EEA). However, authentication can be a huge source of friction for an online customer, potentially leading to a drop-off, in which case the purchase is not completed.
There are many different forms of authentication, such as 3DS1, 3DS2, and Touch ID, and different customers and their banks each have their preferences for the form of authentication.
I led the development of a machine learning model in our payment authentication product, which learns to choose and present the optimal form of authentication to the customers to maximize the chance that they complete the purchases and simultaneously maximize the chance that the bank accepts the payment.
This model processed online payments in real time (around 10 million payments per day), resulting in a significant increase in revenue for the company and its clients (millions per year in additional revenue) by increasing the conversion rate of online payments.
There are many different forms of authentication, such as 3DS1, 3DS2, and Touch ID, and different customers and their banks each have their preferences for the form of authentication.
I led the development of a machine learning model in our payment authentication product, which learns to choose and present the optimal form of authentication to the customers to maximize the chance that they complete the purchases and simultaneously maximize the chance that the bank accepts the payment.
This model processed online payments in real time (around 10 million payments per day), resulting in a significant increase in revenue for the company and its clients (millions per year in additional revenue) by increasing the conversion rate of online payments.
Real Time Dynamic Pricing of Financial Assets within Finance Company
I developed and implemented a dynamic pricing (rule-based and machine learning) model for financial assets within a consumer-facing application. I iteratively improved upon the model by designing AB tests/experiments, ultimately leading to a 20% increase in revenue, around ±10M USD per year.
Education
2015 - 2018
Master's Degree in Theoretical Physics
Leiden University - Leiden, Netherlands
Skills
Languages
Python, SQL
Paradigms
Data Science, Azure DevOps
Other
Machine Learning, Statistics, Bayesian Inference & Modeling, Bayesian Statistics, Big Data, Experimental Design, A/B Testing, Artificial Intelligence (AI), Generalized Linear Model (GLM), CI/CD Pipelines, K-means Clustering, Mathematics, Markov Chain Monte Carlo (MCMC) Algorithms
Libraries/APIs
PySpark
Platforms
Visual Studio Code (VS Code), Docker
Frameworks
Flask, LightGBM
Tools
Apache Airflow, Azure ML Studio, Tableau, Amazon Athena
Storage
Amazon S3 (AWS S3)
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring