Business segment first background

A boutique of tailor-made data science solutions

Data science can provide value to every business, but competent data scientists are rare. We make state-of-the-art research available to anyone.

Our Mission

Empower businesses of any size with a rigorous and effective approach to Data Science.
Check

We do

  • - Machine Learning on tabular data
  • - Time Series Forecasting
  • - Mathematical Optimization
Forbidden

We do not do

  • - Dashboards on Qlik, Power BI, Tableau
  • - Computer vision
  • - Natural Language Processing

= Our favourite technologies =

PythonPytorchGoogle Cloud

Why us?

QualityFluo circle
Quality
data science is not easy, incompetence is expensive. We value quality at every step, from literature review to production deployment.
MethodFluo circle
Method
data science is… science. Blind experiments do not provide insight and business value. Following a rigorous method does.
ResearchFluo circle
Research
data science is evolving fast. We review scientific literature before any project, and work with universities to connect research and application.

Data science

Our Method

This picture is just a model. Wrong, yet useful.
A synthesis between our experience, CRISP-DM and Microsoft TSDP.
Left arrowRight arrow
1
2
3
4
5
Green arrow
Business Understanding
Formally state the problem, choosing suitable evaluation metrics. Review literature for state of the art solutions.
Left arrowRight arrow

Let's work together

Anonymous

You

  • - are expert in your business domain
  • - set goals and constraints
  • - provide constant feedback
  • - are an active member of the team
Curved lineLine

We

  • - are experts in Data Science
  • - propose a way to reach the goals
  • - periodically share with you insights, results, issues
  • - help you interpret the findings of our analysis and get value from the results of our models
We foreground
You background
Home background feedback

Collaboration

Different approaches for different needs

Service Oriented

  • Prerequisites: a problem to solve with data.
  • Sometimes final goal and required data are not known in advance. We can help you with that.
  • We will plan together iterations, each having an agreed scope and duration. At the end of an iteration we decide together whether the final goal is reached and whether it is worth continuing or not.

Product Oriented

  • Prerequisites: clear scope, goal, and data.
  • After a preliminary study, scope of work, economic budget, and plan are agreed.
  • Still, we will work side by side with you throughout the whole project.
  • What if new requirements emerge? If budget and plan do not change, they are accepted in the scope. Otherwise, new budget and plan are discussed.

Get in touch!