Phytorion Fully Custom Solutions
We have been involved in many projects where the customer requirements are so unique that it is better to build a custom data warehouse and analytics solution rather than use an existing data warehouse as a starting point. In these projects, spanning corporate, higher education and government segments, we have designed and deployed custom data warehouse and analytics solutions involving student administration, HRMS and finance applications.
- Build a Kimball-based data warehouse and business intelligence interface (dashboards, cubes, reports) which observe precisely the organization’s processes and data configuration
- Conduct the work in close collaboration with your representatives using a continual feedback loop
- Follow an agile methodology where appropriate to create quick successes. Release data warehouse components as they become ready to allow users to understand how these data structures will answer their business questions, and incorporate their feedback back into the data warehouse
- Keep close focus on Knowledge Transfer
- Take a strong interest in the long-term success of the data warehouse and do the work necessary to accomplish it (this is the work that is not specified in a contract but only becomes evident as the work progresses).
There are several activities in our methodology. We start our work with a Readiness Assessment that includes any Master Data Management or Data Governance needs you might have. We discuss the overall approach to the project and we plan our initial interactions with your constituencies. Our intent is to identify the business areas ready to participate along with their high-level requirements and BI goals.
Once we have established the larger context of what needs to be accomplished now versus later, we plan our project incorporating agile principles. Our aim is to (a) build a comprehensive and flexible architecture, (b) create quick successes by releasing fully working software early and often and, (c) have the users become fully invested in the project's long-term success.
Within the framework of the agreed-to scope, our data architects start working closely with each user team to go through their analytical requirements in detail. This activity is remarkable for the way in which it brings – usually for the first time – an enterprise-wide perspective to requirements that until then were focused on departmental or office needs. Users come to these sessions with a few requirements in mind and leave having seen how others do similar things perhaps somewhat differently. The resulting requirements are more than the sum of the individual needs the users had in mind before they came to the sessions – they are requirements that are better informed by the collective discussion.
The challenge in this foundational requirements activity is the resolution of the rules that users from different offices have observed in prior years. Elevating – and possibly modifying – an office rule to make it work for the department or the enterprise is a process that we manage carefully to resolution, a process critical to the success of the project.
Each functional component or agile iteration is moved to design, paying particular attention to conformed dimensions and facts – the connection points that allow you to extend the data warehouse in the future as your needs dictate. We code using the ETL tool of your choice and perform detailed testing. Coders check for syntax errors, missing values, invalid look ups, etc. and our architects ensure that what was specified is what was coded, whether it is expected values, data warehouse transformations, or historical snapshots.
- Learn the new data warehouse structures
- Learn the new data
- Stress test these structures to ensure they precisely fit their needs
- Learn the Business Intelligence tool
- Gain a full understanding of both the power and the potential of their solution
- Become and remain engaged in the project’s development and deployment
- Communicate to the rest of the user teams the process and the results.
Once users have had an opportunity to learn the new tools and data, we start releasing dashboards, cubes and reports, objects that will constitute the new information environment.
Given our approach, Knowledge Transfer is no longer an isolated activity but an integral part of the implementation – something that starts with the business requirements sessions. In addition and together with you, we plan a number of activities to ensure your team is capable and comfortable assuming ownership of the data warehouse and its analytics once the Phytorion team has completed its work. These scheduled activities may include functional overviews where we review the data warehouse's content, and technical and operational overviews for the IT team responsible for maintaining it. It is important to emphasize that if any of your staff want additional help that goes beyond the scheduled sessions, we are more than happy to meet with them as necessary.
We also work with you to design the training program that suits your organization. Whether it's train-the-trainer, or a wider approach, we focus on the three key elements: training on the tools (ETL, BI), training on the data ("what does the data that I see mean?") and training on the objects delivered (dashboards, cubes, reports, etc.). We then create major user categories, determine their needs, and create the programs appropriate for each category. So where, for example, the core data warehouse team may require extensive training in all aspects of the solution, business analysts may require an overview with emphasis only on one area.
Finally, after Go-Live, we continue to be available to provide ongoing support.