(US & Canada) VIDEO | Common Data Language Is Critical to Data Strategy — Kraus-Anderson CIO
Tony Peleska, CIO and Head of Digital Transformation at Kraus-Anderson, speaks with Thiago Da Costa, CEO of Toric, in a video interview about the data scenario in the construction industry, the organizational data strategy, his 80-15-5 approach to transformation, setting a common data language, educating the organization on data terms, building data pipelines, and creating a data pond to aid digital transformation.
Founded in 1897, Kraus-Anderson Construction is among the top 20 construction firms in the U.S. Midwest. Toric specializes in data movement and data ingestion pipelines for all major construction data sources.
Peleska joined Kraus Anderson as a CIO at the break of the pandemic in January 2020. He supports the development of the financial services, and realty and construction arm of the company. He provides support to the insurance sector of the organization as well.
When asked about the organizational data strategy, Peleska reveals that he was hired to lay a good data foundation which is critical to digital transformation. Further, he shares that after entering the construction industry, he found that there are many systems built to be the source of truth for construction. However, these systems did not have a common interface.
A part of the data strategy lies in defining systems and sources of truth first, agnostic of the goal, Peleska says. The emphasis would lie in defining the current state and ensuring what source of truth means for each group of people and their roles within the organization.
Furthermore, Peleska discusses using the 80-15-5 rule to define the current state, wherein, 80% is about people and how they use tools, 15% is about processes, and 5% is about technology. When it comes to data strategy, he notes that it is critical to have a strong handle around people and processes, and set a common data language within the organization.
Explaining further, Peleska states that in the construction industry, different groups speak differently across markets. Therefore, a massive part of the data strategy is understanding the differences and defining the source of truth to get to a standard level of talking about things, he affirms.
In continuation, Peleska says that it is fundamental to define the system of records used to create sources of truth and know what the pipelines look like. He maintains that it is crucial to understand if there are the right people and roles to define data strategy.
Through the journey towards maturation, Peleska found that there were rampant custom data fields across the organization. This called for the need to look into data modeling and decision-making for the scope of change and custom data fields, he says.
In addition, data was also being stored at different time intervals in different systems, which made data inaccurate, informs Peleska. He states that it is normal to have data that becomes the source of truth at different intervals, while it transitions through its life cycle.
However, Peleska mentions that even after the transition, some groups were still looking at the old data as the source of truth, and such discoveries were made through the journey.
Moreover, the core systems lacked definition, therefore, he initiated talks on data processing and warehousing. Peleska maintains that since there was no data strategy earlier, data was only stored in and used from systems, and there was no external view of how data could be used for analytics, modeling, and Power BI. He mentions that he had to set the stage and start educating on these factors
When asked about data pipelines, Peleska mentions creating a data pond instead of a data lake, to start small and do some proof of concepts (POCs) for people to see and understand. For data pipelines, the organization focused on defining core systems, including Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), and financial systems for budgeting and planning.
Peleska notes that these are the core systems out of Extract-Transform-Load (ETL) that have data that needs to be leveraged for future analytic usage. He took these core systems and created a data pond. Speaking of tools, he mentions using Pentaho, an open-source software, as the core ETL tool.
Contrary to his previous roles, Peleska had to transform the roles and responsibilities of people after coming into the construction industry. While hiring, he had to get a database administrator DBA in-house who understood data modeling enough to start a conversation and help create a data structure.
To create a successful data structure, Peleska communicated with the organization as a whole to understand their take on what they wanted to do to bring change. He asserts that it went from having proof of concepts (POCs) to a minimum viable product (MVP) to having a structure and a foundation.
In conclusion, Peleska states that while creating the pipeline structures to produce the desired output, the organization leveraged the capabilities of Pentaho as its main ETL tool. Thereafter, the organization was introduced to tools like Power BI, that enabled the users to get into the analytical world to see dashboards around historic data stored in systems.
CDO Magazine appreciates Tony Peleska for sharing his insights and data success stories with our global community.