top of page

Nestor > Phase 2



The goal of this phase is to find out if the project idea is actually feasible with machine learning technology. This means that the required data must first be purchased and the selected machine learning algorithm must be trained with the brought data. Thus, this phase is the most crucial for the success of the project. It is therefore important to make the client understand that no promises of success can be made until this phase. If feasibility cannot be guaranteed for reasons such as the data lack of qualitative and/or quantity, phase 1 must be performed again. In that phase, new product solutions are designed without machine learning technology. 


After the second budget proposal is accepted by the client, the Machine Learning Engineers buy the data sets needed for the project. After that, the second phase can be started. This phase of the project is divided into two steps, which must be carried out one after the other. 

First, the Machine Learning Engineers (usually there are two, one junior and one senior) create a Proof of Concept (PoC) with the purchased data sets. For the Machine Learning Engineers, this procedure is already a kind of prototyping. After cleaning the purchased data, an algorithm is defined and the machine learning model will be trained and evaluated with the data. 

This step is very critical and determine the success of the project. If sufficient performance quality cannot be achieved, a new solution must be developed without the technology of machine learning. Therefore Phase 1 of the process model must be conducted again. The performance depends on the quality of the data available. Only when the performance achieves sufficient quality of the desired results, the second step can be carried out. It is very important that the client knows exactly that the foundation for the next steps is established in this phase. 

Concept Creation-01.png


After the second and most critical phase is completed and the Proof of Concept has proven that the desired product performs sufficient with machine learning, a new budget proposal is created. This allows a more detailed project effort to be estimated and thus it is recommended that a budget proposal combining the third and fourth phases be prepared. When preparing the third offer, it continues to be important that all key persons are present and that the offer is reviewed by others who are not involved in the project. 

Share your experience with me

Since project «Nestor» is an ongoing process there are still many things that need improvement. Every feedback on the content, layout, or display of the information is useful. Questions of comprehension are also welcome or share your experience with me. Every constructive feedback helps me to make my practical master project better.


Your help and experience matters.

bottom of page