Data Analytics Portfolio Project
The first stage is all about “What do we want?” Project planning is a vital role in the analytics delivery lifecycle since this is the part where the team estimates the cost and defines the requirements of the project. This stage answer key questions from the student:
- What data set and source will they be using? If the data set is public, then the student should provide a link to the data set. If the data set is not public, then the student should provide a summary of the observations, variables and format they have them available.
- What is the core question the analysis will seek to answer?
- How will we measure the results once they are presented?
The second step is gathering maximum information from the client requirements for the project. Discuss each detail and specification of the project with the customer. The team will then analyse the requirements keeping the design and architecture in mind. The main goal of this stage is that everyone understands even the minute detail of the requirement. Databases, datasets, integration requirements, platform, dimensions, key metrics are to name the few requirements. The key questions to be answered in this stage will be:
- What software requirements does the student have?
- Is there any integration between the data set and the software that needs to be well defined?
- Are there specific timelines or deadlines that need to be met within the logic of the project?
- What Data or DevOps platform will be used to share updates about the project?
- What language or tool will be used to generate the data product?
- Are there any risks involved in the the deployment of the products or the analysis?
- How can we mitigate those risks?
In the design phase, the team scrutinizes whether the prepared design fulfils all the requirements of the end-user. Additionally, if the project is feasible for the customer technologically, practically, and financially. The key questions to be answered in this stage will be:
- Has the student met with the client to determine their needs?
- How will success be defined in the context of this project?
- Do we have any budget constraints that need to be taken into consideration for the project?
- Are there are workload/implementation risks we can envision?
- How can we mitigate the workload/implementation risks?
Time to code! It means translating the design into databases, data lakes, and datasets. In this last stage, the tasks are divided into modules or units and assigned to various team members. Each team member will then start building the system by creating the databases, aggregating and cleaning the data, creating the analysis tools, reports, and dashboards.
This process will have several stages:
- Set up and access to all the sources
- Initial cleaning and integration of the data
- Development of descriptive and summary statistics for the product/model
- Development of an initial draft of the data product
- Validation with client of the requirements and structure of the data product
- Adjustment of data product based on feedback from client
Once the developers finish the build, then it is deployed in the testing environment. Then the testing team tests the functionality of the entire system. In this final phase, the testing is done to ensure that the entire application works according to the customer requirements
The key stages for this section are the following:
- Pilot testing of the data product
- Deployment of the data product
- Validation from client and feedback
- Evaluation of the student and the data project