Data Analytics

Data science and data analytics are two of the pillars of digital transformation. They fundamentally change established practices of how companies organize, operate, and create value. Data-driven business decisions give a competitive edge to organizations by reducing executive risk and increasing operational efficiency. Therefore an increasing number of companies put digital transformation as a priority into their business strategies. However, enterprises often lack internal resources and capabilities to analyze data from sensors, ERP systems, markets, etc. and thereby are missing out on opportunities. Deriving the hidden value from your data is our key objective as your data science and analytics partner.

Data Exploration and Strategy

Data science and analytics projects typically have a high-risk profile caused by a large number of unknowns. To ensure the success of our projects and reduce the risk for the customer, our data analysts first review open business questions and the specific use case. These are then associated with the data available and required, which creates clarity regarding project data requirements and sets the stage for data exploration. In this phase we perform the exploratory data analysis, getting to know the data in more detail and evaluating the feasibility of potential solutions. As a byproduct, additional potentials and valuable insights for the stakeholders are identified. Together with customers we then create a strategy to approach the defined business problems, based on feasibility, risk, merit order, and dependencies. The preliminary analysis output helps business and technical stakeholders better understand the problems, opportunities, and tasks at hand. This ensures transparent communication and efficient project development during the later stages.

Analytics Protocol

To guarantee that our data science and analytics services are reliable, we have agreed upon an internal process that begins with the setting of a use-case-grounded protocol. This protocol defines:

  1. Goodness measures for the specific use case
  2. The evaluation steps ensuring no data leakage or (un-)intentional tampering with results
  3. Specific acceptance criteria for the results
  4. Termination criteria to minimize the risk for our customers

Following this rigorous approach, our team ensures that defined requirements get met. The business implications are results that can be confidently recommended and acted upon, without the risk that our models are giving falsely interpreted results based on unreal quality metrics.

Data Analysis and Solution Generation

This part of the process generates the most visible results to our customers and includes the following sequential steps:

Data analytics process diagram.

Build, train, and evaluate are decisive for generating ideas, building models, training, evaluating, and improving them until the required quality is reached. The incremental process allows for efficient communication between our data scientists and the relevant business stakeholders. Moreover, the iterative process paves the way to an efficient collaboration between our data science team and the domain experts, allowing us to present and discuss discrete findings and results more often. The final evaluation step allows for unbiased and rigorous evaluation of the solution with fresh data. Once the final evaluation is passed and the solution is accepted, we focus on an efficient way of communicating the results, that generates actionable recommendations to the business stakeholders.


Reasonance aims to create data science and analytics solutions that go beyond single-time use and could become part of a continuous value generation chain. Starting with this goal in mind, our team creates and follows a well-defined development strategy covering all project steps from the prototype in the form of an MVP to the fully operational product or service. This guarantees that our solutions could be taken into production without additional unnecessary conceptual efforts and integrated into the existing processes and environments of the customer. Depending on the specific project requirements, our solutions could be built as standalone web or desktop applications or be integrated into the customer’s IT ecosystem with multiple options for cloud deployment. For more detailed information, please contact us or visit our software and cloud engineering service pages.

Interested? Get in touch with us.