Data Scientist

Data Scientist
Make the industry more sustainable and help write our company’s future

The Challenge
Data Science is at the heart of what we do at Sensorfact. Our growing customer base generates 20 million energy measurements per day from a variety of industries and types of industrial machinery. In order to scale and improve our services, we are finding innovative ways to automate and personalize the savings analysis that our energy consultants conduct. But most of all, it involves discovering new data-driven opportunities to reduce waste and optimize savings using machine learning. As a Data Scientist you will be pushing the boundaries of what can be achieved in a production environment with time series forecasting, pattern recognition, anomaly detection, active learning and even causal inference. Your work will be crucial to our company’s future: the algorithms you create create value out of data.

What you’ll be doing

• You will be responsible for designing machine learning experiments to tackle use cases with energy data such as peak shaving, consumption forecasting, machine state detection, predictive maintenance.
• You will work closely with our savings engineers and energy consultants to design and automate new algorithms to support their savings analyses.
• You will work closely with our Data Engineers to create deployable artifacts that will continuously generate insights across our customer base in a mix of streaming and batch applications.
• We have the ambition to become a best-in-class data science team. You will be able to work with all the state of the art machine learning (cloud) tools that you need to do the job, and pick up engineering skills along the way. We further encourage you to hone your skills through personal coaching and training.
• Being part of a scale-up, you are proactive in prioritizing and solving the needs of our fast growing group of customers.

The key technologies you will be working with
As we are scaling up our platform with a small team, we leverage new technologies to keep performance and productivity. Data is accessible through GraphQL APIs managed by Hasura. Time series data is stored in InfluxDB and Postgres and MongoDB for everything else. Data analysis code is Python. We use Jupyterhub to experiment and interact with analytics models and present them to energy consultants. We schedule our distributed processing pipelines with Prefect and are planning to move towards serverless and event-driven architectures. Our source code is on GitLab and we use a mix of GitLab CI and Jenkins for CI/CD.

How we do it
We do Scrum with 2-week sprints, sprint planning and retrospective sessions. Our stand-ups are at 9:45 and if you're not there you can chime in over Meet. We keep track of things using Linear, Google drive and Outline, and we stay in touch with each other over Slack. We work closely with the product department to collect new ideas to work on from business stakeholders and customer feedback.

We know how important it is to get in the zone and write beautiful code so we schedule most meetings in the morning and keep the afternoon quiet (we try). We work from home about 70% of the time, but we enjoy meeting each other in the office regularly -- covid allowing of course.

You are perfect for this job, because you…

• Have an MSc or PhD in Artificial Intelligence, Data Science, Econometrics, Computer Science, or a comparable analytical / technical field
• Are a medior (3+ years) all-round data scientist who is fluent in data analysis with Python, e.g. with Pandas and numpy
• You have a solid background in statistics and have experience with machine learning tools such as scikit-learn, Tensorflow and/or Pytorch
• Are passionate about at least one of the following: time series forecasting, pattern recognition, anomaly detection, active learning, causal inference
• Have experience in integrating algorithms into existing production systems, or a burning desire to learn it
• Are familiar with standard software engineering practices such as version control, testing and deployment
• Business level fluency in English is mandatory
• (Bonus) have experience with distributed processing frameworks such as Spark or Dask

What we offer
A full time position (32-40hrs), salary, pension, working from home, team activities, training budget – the usual. We work in a forward-thinking start-up culture with an energetic and engaged team, located around the corner of Utrecht Centraal. We’ll provide you with an NS-business card or cover your travel expenses to get there. We know how important it is to have the right tools. Any hardware or software you need to get your job done: great monitor, the best laptop, standing desk – you’ve got it.