Image: Wonderlane


Overview of our services:

For further Data Science services like Advanced Analytics Consulting, Model Review, Project Management, Coaching, Training, Presentations, etc... please contact us. We will be pleased to assist you.

Get in touchLectures

Potential Assessment

Are you sitting on a mountain of data and don't know how to mine the supposed gold in it? Usually asking the right questions is the biggest hurdle. We are happy to assess the potential of your data and evaluate the best approach with the greatest chance of success.

In our Potential Assessment we examine the following points:

  • Business Objectives: Which goals are you pursuing? Which problems are you trying to solve? What hinders you in the implementation?
  • Success Metrics: How should the success of a possible solution be measured?
  • Data Landscape: What data is available? In which quality? What data is missing?
  • Tool Landscape: What analytical tools are in use already? What new tools would fit into your IT infrastructure?
  • Skills and Resources: How is the internal level of knowledge regarding advanced analytics? Which existing resources can be used? What stakeholders must be convinced yet?
  • Potential for Success: Which profit or what savings can be realistically expected?

The scope of a Potential Assessment varies depending on the complexity of your goals. We can help you starting with non-binding meeting up to an agile pilot project. We are looking forward to making you a tailor-made offer according to your needs.

Get in touch

Data Preparation

Real-world data is often incomplete, noisy and inconsistent. The phrase "garbage in, garbage out" is particularly applicable to machine learning projects. Before a software algorithm can go looking for answers, the data must be cleaned up and converted into a unified form that machine learning algorithms can understand.

Major tasks in data preparation are:

  • Data Consolidation: Collect, select and integrate data from multiple data sources
  • Data Cleaning: Fill in missing values, reduce noise in data, identify or remove outliers, and resolve inconsistencies
  • Data Transformation: Normalize data, discretize or aggregate data, and construct new features
  • Data Reduction: Reduce number of variables, reduce number of classes, and balance skewed data

We are happy to help you conducting these essential steps to bring your data science project onto the right track.

Get in touch

Exploratory Data Analysis

Exploratory Data Analysis (EDA) is an efficient way to gain understanding of the main characteristics of your data. EDA is based on numerical and visual methods. Furthermore EDA enables a data driven hypothesis generation. This process often feels like "detective work": We explore your data looking for hidden patterns and structure that lead to new hypotheses and predictive models.

The main steps conducted in Exploratory Data Analysis are:

  • Visualize: Visual methods are applied to "see" underlying structures, detect outliers and identify anomalies.
  • Understand: By asking questions and conducting interactive experiments we deepen our insight into your data.
  • Explain: Whatever results we find in your data we will report back to you in a visual and easy understandable way.
  • Challenge: We are always skeptical about any results and are happy to face the critical questions posed by you.
  • Iterate: Exploring the unknown is never a straight forward process. Iterations are essential in EDA.

Let us help you indentifing patterns in your data and generating business value by exploiting those.

Get in touch

Predictive Analytics

Business metrics do a great job summarizing the past. But if you want to predict how customers will respond in the future, there is only one place to turn — Predictive Analytics (PA). By learning from your abundant historical data, PA delivers something beyond standard business reports and sales forecasts: Actionable predictions for each customer.

Our Predictive Analytics solutions run through the following steps:

  • Project Definition: Define business objectives and the success criteria. Translate the business objectives into predictive analytics problem definition and plan to achieve the objectives.
  • Data Preparation: Collect, characterize, integrate, format, transform and clean data. Create and evaluate features: Identify features most likely to have predictive power. Iterate.
  • Model Building: Select predictive modeling techniques suited for the problem. Generate test design. Build predictive models. Assess models based on predictive performance versus complexity. Select "best" model which best satisfies the requirements.
  • Model Deployment: Implement data preparation process in your environment. Integrate model into your IT infrastructure. Execution of the predictive analytics solution on productive data.
  • Model Monitoring: Monitor performance of the implemented models. Monitor quality of the productive data. Adapt to changing business objectives periodically.

We will enable you to look into the future and make the right decision already today.

Get in touch

Data Visualization

We use Data visualization as modern equivalent of storytelling. The brain processes visual content (like an infographic) 60,000 times faster than it does text. Patterns, trends and correlations can be spotted quickly and easily.

With our interactive data visualizations we go even one step further – moving beyond the display of static graphics to interactive ones. We are letting you drill down into charts and graphs for more details, and interactively (and immediately) changing what data you see and how it is processed.

For creating Data Visualisations we consider the following points:

  • Purpose: Why are we doing this visualization? What is the goal?
  • Content: What are we trying to visualize? What is the story we want to tell?
  • Structure: How are we going to visualize it? As Scatterplot, Treemap, Network, Heatmap,... ?
  • Formatting: How will it look and feel? How will it be consumed?
  • Interactions: How can we engage the viewer? What follow up question can be answered by filtering or drill down?

We love to visualize data. Let us help you to see the insights in your data.

Get in touchExamples