(US & Canada) | Data Governance Is Like Gardening, Keep Pulling Weeds for Life — Centric Consulting Practice Lead for Data and Analytics

Tim Hoolihan, Practice Lead for Data and Analytics at Centric Consulting, speaks with Robert Lutton, VP, Sandhill Consultants, and CDO Magazine Editorial Board Vice Chair, in a video interview about the challenges in implementing a data governance framework, his role in managing high-performing teams, and leveraging AI as a team to drive innovative solutions.

Centric Consulting combines business and technology consulting experience and flexibility to create tailored solutions for your problems.

The challenges while implementing data governance frameworks arise due to miscommunications, unclear roles, and unclear data ownership, says Hoolihan. The key to data governance is knowing that it is much like gardening, where one has to keep pulling weeds for a lifetime.

The other aspect of overcoming the data governance challenge lies in data ownership, says Hoolihan. He maintains that IT is associated with owning data because data goes into technical tools or data catalogs, whereas business needs to own it.

Adding on, Hoolihan states that it is unimaginable that a person working on the spreadsheets in the database owns the data and defines its importance. The real owners, he adds, are the lead analysts, advisors, and coaches who interpret the data, produce insights, and derive value from it.

Further, Hoolihan states the importance of understanding the roles. It is critical to know where IT can play a nuanced role in tools and also drive crucial questions to understand how the same data can be looked at differently from varied perspectives.

Illustrating with an example, Hoolihan mentions working with an organization wherein product and program managers make budgetary predictions on a yearly, quarterly, and monthly basis. In a certain case, to be updated on project completion, he relies on the most recent prediction. Whereas, when the question is about the best estimator among all, Hoolihan needs quarterly or annual estimates. He mentions that the data and analytics team can help drive questions, and how to look at the data depends on the question in the use case.

Next, Hoolihan comments on building and nurturing a high-performing team for data analytics projects. He says that a high-performing team and technologists are people who are passionate about their jobs. An effective approach with these people, according to him, is servant leadership, or being the customer service for team experts.

Looking back on the projects Hoolihan managed, he notes that data scientists’ knowledge of advanced AI, for example, far exceeds his own. His job is to ensure that they are applying the knowledge in the right places and to support them when in need.

Commenting on the evolving skillsets, Hoolihan states that, unlike in the past, tool sets now empower people. He stresses the importance of experiment design and good statistical principles.

Business analysts nowadays can be data-savvy and build interesting solutions with the tools, but they need to know the right data to feed in, asserts Hoolihan. Skillsets for him are not advanced math but common sense.

To explain, Hoolihan uses a sports metaphor and says that a model trained for NFL players should not be applied to high school footballers because the attributes are completely different.

Moving forward, he applauds the analytics group at Centric for bringing in the best of both breeds, with knowledge of traditional, structured data while also adopting data lakes. To drive innovative solutions, the team leverages AI to do forecasting, clustering, categorization, and so on within the system.

With traditional statistical methods and regression techniques, Centric Consulting has been making predictions for a long time, says Hoolihan. The company has become a part of Databricks, acquired the Azure skill set, and is looking into more deep learning in Python on top of Spark.

Sharing an instance, Hoolihan states that a situation arose where a group of estimated program and project managers had to fill out an intake form for project and technical architects. The project estimation was done using a prediction model on Databricks.

The small dataset size meant accuracy was not optimal, as larger training sets are preferred in machine learning. However, this rough estimation still proved valuable, helping architects significantly in their tasks, says Hoolihan. Considering the high demand for technical architects, the company benefited greatly from this approach, saving both time and resources, he notes.

Furthermore, the company is also looking at the advanced breakdown structure, licensing costs, and returns. Also, there have been ongoing conversations around AI usage with organizations working with Centric Consulting.

In conclusion, Hoolihan states that the company tries to understand the organizational vision and problem areas. Then it assesses the governance structure and assigns the right people to carry out the vision work holistically, ensuring safe AI usage.

CDO Magazine appreciates Tim Hoolihan for sharing his insights with our global community.