Who are we?
Denexus' mission is to build the global standard for Cyber Risk Quantification and unlocking the Risk Capital Markets to underwrite cyber risk at scale.
Join us in developing the world's first statistical engine for measuring industrial cyber risk.
Duties and responsibilities
As a data scientist, you love creating statistical models and are passionate about making sure that your models accurately describe the real world by making sure that they preserve qualitative properties of the observed data. You aren’t scared to build a new model from scratch according to specifications provided by subject matter experts, find thrilling the challenge of facing completely new problems every day and having to find ways to tackle them, and love being in constant learning.
The ideal candidate has solid prior background in statistical modeling, having worked as a data scientist using techniques that create explainable models like Bayesian modeling instead of black box models like neural networks.
· Decide which models and statistical distributions more accurately describe frequency and impact of cyber attack events
· Create models that give the full probability distribution of outcome events, rather than just a point estimate of the most likely outcome
· Test said models, find corner cases for which they don’t work accurately and figure out ways to improve them
· Help an interdisciplinary team define the best way to build cyber attack chains from standalone alerts. For this task, use of more traditional black box machine learning techniques might also be useful
· 2+ years experience with statistical modeling
· Familiarity with Bayesian modelling is a must
· Fluency programming is a must (you will be mainly using Python)
· Experience with machine learning (Neural networks, boosted trees, SVMs) is a plus
· Experience with clustering machine learning techniques is a plus
· Remote working in Europe. Most of the team you will be working with is based in Madrid, Spain