loader

Introduction

In the first part of our series on Data Governance, we looked into what is data governance, why we need it and what are the future trends. 

In the second part, we discussed various Data Governance frameworks used by organisations to implement data governance. 

In part 3, we will discuss various Data Governance operating models and services required by corporations to fulfil the same. 

Organisations choose one of the three operating models for data governance: 

  1. Centralised 
  2. Decentralized 
  3. Federated 

While each has its own pros and cons, no one model is superior to the others. The choice is often left to the organisation based on practicality of application.  

Centralized Operating Model

This model is a top-down model that relies on a single individual (typically Chief Data Officer or Chief Information Officer or a dedicated Data Governance Director) to make key data governance policy decisions and provide directions for the organisation. 

This model leads to rapid decision making and clear-cut data governance structure. Organisations, specially who are low on data governance maturity, can get started very easily and quickly.  

However, such simplicity and centralisation lead to operational rigidity in longer run. This may not be feasible if the organisation is spread across multiple jurisdictions with differing regulatory compliances.

Decentralized Operating Model

In this model, there is no single owner, and ever decision is taken by committees. This leads to relatively flat data governance structure. Each committee is aligned a particular line of business and have deep domain knowledge that they can bring to the table. 

However, co-ordination and consensus can be a challenge as each committee takes a decision that is based suited for the LOB in question. This can lead to differences in policies between various LOBs. 

Hybrid/Federated Operating Model

This model tries to capture best features of the previous models. A centralised governance body oversees data governance at the enterprise level and provides the framework, tools, and best practices for LOBs to follow. The LOBs have authority to implement data governance in a way that best suits their operations and data practices.  

Such centralised data governance strategy with a decentralised execution is well-suited for transnational organisations, which are bound different and sometimes contradictory regulations across different geographies. This makes the data governance policies very user-oriented and iterative leading to development of innovative data products. 

However, not all organisations can implement such operating models as they may not have the sufficient data governance maturity to initiate. Oversight is very essential to reduce discrepancies between business units (like, metadata management, data glossary, etc.) and maintain balance between priorities of the organisation and those of the individual BUs. 

Data Governance Services

  • Data Governance Strategy — Analyse the data governance maturity of an organisation through discussion with key stakeholders, identify gaps in existing system, define the organisational goals, process, best practices, and technology. 
  • Data Pipeline and Data Lake Build — Develop a single source of truth by bringing siloed data (structured, semi-structured and unstructured) across various systems and applications into a centralised data lake. 
  • Master Data Management — Build a consistent and reliable record of data (derived from different internal and external sources and applications) through data de-duplication, reconciliation, and enrichment processes. 
  • Data Catalogue and Discovery — Categorisation of data assets in a systemic and automated manner to make it searchable and discoverable by data engineers and consumers. 
  • Data Classification — Tagging data with appropriate metadata to provide better understanding of its content and context to secured usage. 
  • Data Security — Protect the data both at rest and in motion through encryption, masking, and tokenisation. 
  • Data Privacy and Sharing — Create the right framework to ensure data is stored, accessed, processed, and shared as per relevant privacy regulations. 
  • Data Quality — Ensure data is accurate, complete, consistent, and valid so that data engineers and consumers can readily use them. 
  • Data Lineage — Create process to identify the origin of data, how was it processed, who accessed it, and where it is being used. 
  • Data Stewardship — Build an overarching framework to guide data producers, data engineers and data consumers on data is owned, processed, utilised and accessed. 

Translab Technology DGCOE

  • Translab Technologies has developed a Data Governance Centre of Excellence (DGCOE) to work with organisations deluged by data and data complications and provide both strategic guidance and tactical execution that alleviates such challenges. We will undertake a three-phase approach to address such needs. 
  • In the discovery phase, our interdisciplinary team (consisting of technical and domain experts) will identify gaps in current data governance maturity level with that needed to meet the business goals and develop a roadmap. 
  • In the execution phase, we will work with the business units to implement the roadmap. Our team members will be deployed at your locations and work alongside your teams to build and deploy the detailed plan. We will customise the plan as required by the LOB, application, region or operational model 
  • In the maintenance phase, we will ensure continued compliance with the data governance framework and regular updates as per business and regulatory changes. 

Reach out to us for your data governance need by filling in the contact form. 

Leave a Reply

Your email address will not be published. Required fields are marked *