Anniek Hegeman Email & Phone Number

Utrecht, Nederland

Utrecht, Nederland

Utrecht, Nederland

Isatis Business Solutions (IBS) is een partner voor dataoplossingen dat actief is in vele sectoren. Het Data Science team binnen IBS deelt kennis met elkaar om altijd up-to-date te blijven van de laatste data science gerelateerde ontwikkelingen. Hierin worden wekelijkse presentaties gegeven.

Mosadex E-Health makes innovative software solutions to contribute to the optimization of patient care in the Netherlands. Together with its sister companies, they support pharmacies with innovative pharmacy concepts. One of the sister companies is Mosadex Central Filling. Mosadex Central Filling is a sorting company that sorts medicines from the wholesale center into prescriptions for the pharmacy. This is done by robots, for which the input must be optimized to get the maximum number of recipes from the robot. For this optimization problem I built a Greedy sorting algorithm in Python to determine which products should go into the robot and where, to make the filling process as efficient as possible.Results achieved:- The stock-keeping process in the warehouse recreated in an object-oriented framework in Python.- Implement Greedy sorting algorithm.- Achieved an improvement of €1000,000+ extra profit per year.- Present project to Mosadex E-Health wide Data Team.- Code and data processed via Azure.- Regular communication between stakeholders, drafting requirements, translating from business to technical and technical to business.

Amsterdam En Omgeving, Nederland

Within Bright Cape I have worked on several projects. VIPO | Image Recognition I have worked as a data scientist for deep learning project aiming to operationalize image recognition algorithm for fashion recommendation. Specific imaging tasks include segmentation and classification with the aim to detect gender, age category and fashion attributes of target images. As a data scientist I was responsible for the development of classification algorithms by collaborating with other data scientists.Techniques used in the framework involve amongst others: Python, Jira, GitHub, AWS, Jira, Agile, Tensorflow, Neural Networks, Image Recognition, Segmentation.Dronesurance | Development Risk EngineDronesurance is an EIT-funded innovation project to develop a risk profile tool for insurers to determine mass market insurance premiums, where the focus is pay-as-you-fly real-time episodic risk insurance for drones. As a data scientist within the project it was my resposibility, amongst other data scientists within the team, to develop a risk engine to determine a fair premium for each unique drone flight.Techniques used in the framework involve amongst others: Python, GitHub, Jira, Agile, Bayesian Logistic Regression.WALDO | A Universal Framework for Anomaly DetectionWALDO is an EIT-funded innovation project, focused on developing an open-source anomaly detection framework that consolides specialized and independent anomaly detection solutions. The anomaly detection framework enables companies to work safer and efficient, as instances can be monitored continuously without the need of experts. As a data scientist I was responsible to create a generalized Kedro framework which consisted of connected modules which ranges from pre-processing till anomaly detection models by collaborating with other data scientists.

I have worked on an EIT co-funded innovation project, UAV Retina, with other European industry- and research partners. The goal of the UAV-Retine project is to create a flexible autonomous UAV (Unmanned Aerial Vehicle) platform for Search-and-Rescue (SaR) operations. In the project, I worked on the data analytics module for the avalanche rescue operations. Getting to know the domain, gathering the necessary data sets, developing quantitative models and methods for predicting and visualizing the likelihood of the victim's location. Through mathematical analysis of historical data and characteristics of the landscape, the belief map indicates the probability of location to contain a vicitim. This belief map is used in initial phase of the SaR operation - hotspots are used to direct the UAV to the most likely victim's locations. UAVs are a faster, more flexible and less costly alternative compared to helicopters. Techniques used in the framework involve amongst others: Python, R, QGIS, Text Mining, Spatial statistics, Regression models.

To complete my master's degree I did a research on the recovery duration of athletes with a sport-related concussion. I created models using different machine learning methods, like Random Forest and Support Vector Machine, to predict the length of recovery for athletes with a persistent sport-related concussion. In this way, clinicians could give patiens a more individual targeted treatment.

Amsterdam

Together with Babette Mooij, I started the HR-software company Growthrs. We wanted to help fast-growing companies set up their people departments and give everyone at the company insights into their teams, skills, and organization. Babette and I participated in the ACE startup incubator program to further build Growthrs. The project is now on hold indefinitely.

Delft En Omgeving, Nederland

Den Haag En Omgeving, Nederland

Den Haag En Omgeving, Nederland

Delft, Zuid-Holland, Nederland

On the 19th of May 2017 I organized the StuD Careerday 2017. This is a day where over 140 students of the Delft University of Technology can meet companies and attend workshops. Inspiring lectures can be taken and workshops, where the students learn soft skills, can be done.

Master Of Science - Msc, Data Science

Bachelor Of Science - Bsc, Applied Mathematics