Experience

Digital Sundai
• Netherlands
• IT Services and IT Consulting
• 1 - 100 Employee
• Principal ML Engineer

  • Sep 2022 - Present


KPN

• Netherlands
• Telecommunications
• 700 & Above Employee

• Senior ML Engineer

  • Jan 2020 - Jun 2022

• Data Scientist / ML Engineer

  • Oct 2015 - Dec 2019

  At KPN I am responsible for creating end-to-end data-driven business solutions. These projects can encompass the entire data value chain, from extracting data from its source, building a machine learning model to putting the model in production using a self-built, automated pipeline. ProjectsCustomer credit checkCreated an API to perform customer credit checks in real-time using machine learning. Generalised an existing Python code-base to launch multiple machine learning models onto the same platform with the least amount of trouble and refactoring. Wrote an interpreter that converts human readable business rules to Python code. Used Jinja to write a generalised ETL process to extract data from its source and load onto Hadoop file systemModel factoryBuild a pipeline with which colleagues can easily deploy machine learning models onto development and production systems with version control and scheduling. Created an R package to easily integrate transform, store and load model output to a database. Explained colleagues on how to use version control to stay in control of new their models and add new features without breaking production.Churn predictionCreated a random forest model to identify customers that are likely to cancel their subscription in the future. Put model in production where outcome is used by Marketing for targeted campaigns. Updated and tweaked the model to increase accuracy. Explained model outcome interpretation to Marketing to help increase effectiveness of campaigns.Creating and maintaining internal packagesCreated, maintained and extended functionality of various internal packages, including the aforementioned R package for the model factory as well as the corresponding Python package as well as an R package containing various different convenience functions. Show less



Purpose Management Consulting

• Consultant

  • Jan 2015 - Sep 2015

  Used the Microsoft Office suite to build a database and reporting pipeline for the annual mortgage special asset management benchmark Used the Microsoft Office suite to build a database and reporting pipeline for the annual mortgage special asset management benchmark



PwC Nederland

• Netherlands
• Business Consulting and Services
• 700 & Above Employee

• Senior Consultant Data Analytics

  • May 2010 - Dec 2014

  As a consultant in the data analytics department I am working on multiple projects simultaneously, managing junior colleagues and translating the required high-over analysis to detailed hands-on steps to be taken. Projects: - Energy company: Created a consumer usage forecast model to predict annual cash flows and energy tax payable. Achieved 99% accuracy in forecasts. - Malaysian Postal Office: Worked on an international team to implement activity based costing. Automated the construction of all necessary accounts to correctly assign costs to products. - Utility company: Analysed new tariff structure and made recommendations to increase customer focus and transparency. This helped the company in creating effective communication and ultimately increased client acceptance of the new structure. - Packaging sector supervisory foundation: Reviewed third party's statistical analysis of total amount of undeclared packaging produced in The Netherlands. Used a sensitivity analysis and control-round to improve best estimate by 10%. - Internal roles: Teaching the Excel and VBa courses. Developed the VBa course which has been rated by participants with an average 4.6 on a 5-point scale. Show less



Vrije Universiteit Amsterdam

• Higher Education
• 1 - 100 Employee

• Student-Assistent financiële administratie

  • Oct 2005 - May 2006

  Matching incoming payments to outstanding invoices, and creating payment lists for the Faculty of Economic and Business Sciences. Matching incoming payments to outstanding invoices, and creating payment lists for the Faculty of Economic and Business Sciences.