“CIO Perspectives” is a white paper series by Mark Settle that explores top-of-mind technical issues confronting today’s IT leaders. Mark is a seven-time CIO, a three-time CIO 100 award winner, and a two-time book author. His most recent book is Truth from the Valley, A Practical Primer on IT Management for the Next Decade.
Internal Developer Platforms (IDPs) are an essential component of every modern software engineering organization. IDPs offer developers a variety of services that automate common workflows and promote the use of common tools. These so-called golden paths and preferred toolchains are designed to reduce the cognitive workload required to perform routine activities such as establishing working environments, deploying code and monitoring software performance. The ultimate goal of platform engineering is to maximize the time developers spend coding business functionality and minimize the time devoted to all other ancillary activities.
Similar concepts have not been embraced by the distributed data analytics teams found in large enterprises, even though data analysts experience many of the same frustrations as software developers. Data professionals routinely complain about time wasted on a host of ancillary activities that must be performed before they can use their business domain knowledge to cull meaningful insights from corporate data. However, their collective frustration rarely impels large enterprises to establish Internal Data Analytics Platforms (IDAPs) to maximize the productivity of analyst teams scattered across multiple functions or lines of business.
This paper identifies the hurdles that must be overcome to motivate investments in Internal Data Analytics Platforms and it describes the types of services that an IDAP might offer.
Enthusiasm for data mesh organizational principles has stimulated new thinking about ways in which analytics organizations might benefit by emulating selected software engineering practices. Engineering platforms specifically designed to support enterprise-wide data analytic activities patterned after the IDPs found in software organizations would materially benefit the distributed analytics teams found in every large enterprise.
