Mikkel Vilstrup, Developer in Copenhagen, Denmark
Mikkel is available for hire
Hire Mikkel

Mikkel Vilstrup

Verified Expert  in Engineering

Bio

Mikkel is a data scientist and machine learning engineer with solid competencies in both software and business acumen. He recently architected an enterprise search platform based on deep learning bought by several major Danish companies. Mikkel has also led a data science team, introducing new and innovative car insurance to the Danish market. Here he was in charge of both price estimation and risk analysis. In general, Mikkel excels at finding creative software solutions.

Portfolio

Undo Forsikringsagentur A/S
Python, Statistics, Geospatial Data, Machine Learning, PyTorch, Clustering...
Raffle.Ai
Python, Deep Learning, Machine Learning...
Denmarks Technical Universtity
Python, Deep Learning, Amazon Web Services (AWS), Artificial Intelligence (AI)...

Experience

  • APIs - 6 years
  • Python - 6 years
  • SQL - 5 years
  • Data Analysis - 5 years
  • Machine Learning - 4 years
  • Deep Learning - 4 years
  • Data Modeling - 3 years
  • Node.js - 3 years

Availability

Part-time

Preferred Environment

MacOS, Slack, Ubuntu, Email, Cloud Drive, Git, GitHub, JetBrains

The most amazing...

...machine learning model I have created decided the rollout of the 5G network in Denmark.

Work Experience

Lead Data Scientist

2019 - 2021
Undo Forsikringsagentur A/S
  • Researched and implemented new ways to estimate collision risk for car insurance using telemetry data.
  • Analyzed user behavior, pricing, and marketing initiatives to support company growth.
  • Managed and supervised all in-house data analytics and modeling projects.
  • Aligned key stakeholders and led projects aiming to make the company more data-driven.
  • Involved in the entire data pipeline, from data modeling to client visualizations.
  • Developed a custom in-house framework for ETL on top of Airflow.
  • Made a Python to SQL translator to speed up data analysis projects.
  • Created microservices in Node.js and Fastapi for data visualizations.
Technologies: Python, Statistics, Geospatial Data, Machine Learning, PyTorch, Clustering, Apache Airflow, Web Crawlers, Artificial Intelligence (AI), Node.js, React, Data Modeling, Software Design, Software Deployment, Software Design Patterns, Geospatial Analytics, Data Analytics, Jupyter, Time Series Analysis, Amazon Web Services (AWS), GitHub, HTML, CSS, APIs, PostgreSQL, REST, Data Analysis, REST APIs, Microservices, GraphQL, Pytest, Architecture, System Design, Datasets, Django, Docker, Kubernetes, NumPy, Pandas, Data Pipelines

Head of Engineering

2018 - 2019
Raffle.Ai
  • Hired and managed a full team of back-end and front-end developers and ML engineers.
  • Architected an in-house deep learning infrastructure based on Kubernetes.
  • Developed a custom search platform for company intranets based on deep learning.
  • Architected a multilingual text similarity model for text retrieval.
  • Led the development of the client-side Electron application built with React.
  • Created microservices to support the dataflow using Node.js and Python.
Technologies: Python, Deep Learning, Machine Learning, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), Distributed Systems, React, PyTorch, TensorFlow, Artificial Intelligence (AI), Node.js, Data Modeling, Software Design, Software Deployment, Software Design Patterns, Search Algorithm Design, Data Analytics, Jupyter, Time Series Analysis, Amazon Web Services (AWS), GitHub, HTML, CSS, Clustering, APIs, PostgreSQL, REST, Data Analysis, REST APIs, Microservices, GraphQL, Pytest, Redis, Architecture, System Design, Datasets, Django, Docker, Kubernetes, NumPy, Pandas, Data Pipelines

Research Assistant

2017 - 2018
Denmarks Technical Universtity
  • Handled half a terabyte of telemetry data, collected from overweight subjects' smartphones to estimate their activity level.
  • Applied various machine learning approaches including deep learning for human activity recognition.
  • Acted as a teaching assistant for both internal and external deep learning courses.
Technologies: Python, Deep Learning, Amazon Web Services (AWS), Artificial Intelligence (AI), Geospatial Analytics, Geospatial Data, PyTorch, TensorFlow, Data Modeling, Jupyter, Natural Language Processing (NLP), Generative Pre-trained Transformers (GPT), GitHub, HTML, CSS, Clustering, APIs, PostgreSQL, REST, Data Analysis, REST APIs, Pytest, Datasets, NumPy, Pandas

Data Scientist

2017 - 2018
TDC Group
  • Analyzed several Petabytes of network data to find movement patterns and other insights about the Danish population.
  • Worked with Hadoop, PySpark, and Hive on both a test and production cluster.
  • Supervised various in-house projects within NLP, recommendation, clustering, and deep learning.
  • Handled communication with various customers and potential customers.
Technologies: Python, Hadoop, Spark, Jupyter, Geospatial Data, SQL, Apache Hive, PyTorch, Clustering, Artificial Intelligence (AI), Data Modeling, Software Design, Software Deployment, Software Design Patterns, Geospatial Analytics, Data Analytics, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), Amazon Web Services (AWS), Computer Vision, Microsoft Power BI, GitHub, HTML, CSS, APIs, PostgreSQL, REST, Data Analysis, REST APIs, Pytest, Redis, Architecture, System Design, Datasets, Docker, NumPy, Pandas

Software Developer

2015 - 2016
TwentyThree
  • Developed an in-house web crawler that continuously crawled the web for potential leads.
  • Created an integration to HubSpot to provide the sales representatives with continuous leads.
  • Made an in-house model for ranking leads to improve the efficiency of the sales representatives.
Technologies: Python, Web Crawlers, JavaScript, CSS, Jupyter, GitHub, HTML, APIs, PostgreSQL, Datasets

Experience

Fashion Image Search

A custom image search application for fashion images. The application was capable of finding similar clothing images using deep image embeddings combined with metadata. A database with 500,000 images, it could handle 1500 searches per second.

City Utilisation During Yearly Festival

I was in charge of analyzing the movement patterns of all people in the second-largest city in Denmark during the yearly festival using telematics data. The analysis allowed the city to optimize its infrastructure to handle the increased load via a Microsoft Power BI application.

Few-shot Text Similarity Model for Internal Knowledge Bases

A system for generating a base model which could be fine-tuned on new knowledgebases using just a few examples. The system was sold to several companies and used in production for internal and public searches of FAQs and intranets.

Analysis of Car Accidents in Denmark and California

Utilizing public data of car accidents in Denmark and California combined with data from the Open Street Map. I analyzed the risk of car accidents from both a temporal and spatial perspective, showing several similarities across the two continents.

Education

2015 - 2017

Master's Degree in Mathematics and Computer Science

Danish Technical University - Denmark

2012 - 2015

Bachelor's Degree in Software Development

IT University of Copenhagen - Denmark

2010 - 2012

Bachelor's Degree in Psychology

Copenhagen University - Denmark

Skills

Libraries/APIs

Pandas, REST APIs, NumPy, React, TensorFlow, PyTorch, Node.js, Vue

Tools

Pytest, Git, Jupyter, Microsoft Power BI, GitHub, Apache Airflow

Languages

Python, SQL, JavaScript, GraphQL, HTML, CSS

Frameworks

Flask, Django, Hadoop, Spark

Paradigms

REST, Microservices

Storage

PostgreSQL, Data Pipelines, Apache Hive, Redis

Platforms

Software Design Patterns, Amazon Web Services (AWS), Docker, Kubernetes

Other

Deep Learning, Machine Learning, Software Development, Data Analysis, Geospatial Data, Artificial Intelligence (AI), Natural Language Processing (NLP), Datasets, Generative Pre-trained Transformers (GPT), APIs, Data Visualization, Distributed Systems, Web Crawlers, Data Modeling, Full-stack, Architecture, System Design, Software Design, Software Deployment, Algorithms, Search Algorithm Design, Geospatial Analytics, Sensor Fusion, Data Analytics, Time Series Analysis, Statistics, Computer Vision, Clustering

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