In software development, a proof of concept can be a vital tool to demonstrate the software’s capabilities and its fit with the client’s requirements. But how do you go about creating one? A recent example from our data integration work in the Oil & Gas industry illustrates the steps we take to create a successful proof of concept.
You may also be interested in: How to create a successful proof of concept.
Step 1: Defining the client’s requirements
Our clients’ needs can range from testing the suitability of the proposed software to sharing knowledge within the client’s organisation. In our recent Oil & Gas proof of concept, the main aim was to demonstrate the effectiveness of our Transformation Manager data migration tool in meeting the client’s petroleum data management requirements. Specifically, the proof of concept demonstrated that it is possible to use Transformation Manager to efficiently implement and deploy industry data standards to different types of petroleum data files.
In our example, there were two types of files – LAS and DLIS – to migrate in the context of the PPDM and WITSML industry standards. The proof of concept accordingly aimed to deliver three scenarios:
- Embed transforms within near real-time message handling, particularly WITSML
- Data migration converting to LAS files to and from the client’s PPDM-based system
- Data migration converting to DLIS files to and from the client’s PPDM-based system.
Step 2: Defining the client’s input
In any project, we clearly define the requirements from the client before starting work. In this example, the items required before the proof of concept began included:
- Business rules to define the required mappings, such as details of field mappings, lookups and error handling
- Samples of the data to be extracted, such as well header and log curve data
- Any relevant information about the source and target models, such as local choices in the use of PPDM
- Sample files in the required format, including LAS 2.0 and 3.0.
Step 3: Action planning
Our action plan followed a similar path to our previous proofs of concept. It was followed through to implementation and included:
- A review of the information received from the client:
- Each source and target data model, including format, connection options and sample data
- Validation rules
- Mapping rules
- Design of the workflow of the data through the integration process:
- Document flow of data
- Document the decisions made
- Design the integration for implementation within our Transformation Manager software
- Implement the integration within Transformation Manager, including testing
- Demonstrate the results to the client (see Step 4 for more details)
- Describe the process and usage of Transformation Manager to the user
- Create a report describing the concepts, outcome and potential usage within future projects.
Step 4: Proof of concept delivery
In the petroleum data management example mentioned above, our team delivered:
- A final, tested, Transformation Manager deployment pack which delivered the data transformation to the client’s specification
- A licensed version of Transformation Manager for internal evaluation
- A proof of concept report as described above
- Knowledge transfer to the client, including the use of Transformation Manager, a review of its capabilities and an overview methodology
- A presentation and demo showing key processes, outcomes and future options.
The result was a software component which can efficiently deploy the client’s data to a range of the most popular petroleum data management file formats.