Unlock financial services insights in data lineage with Deloitte Data Insights Monitor 

Deloitte Data Insights Monitor offers four sets of capabilities:

1. 360 view on data lineage & insights: See the big picture of your end-to-end data-driven process and track the implementation of data requirements over time
2. Effective business-in-mind implementation: Align your data lineage with your business needs and policies
3.  Versatility in views: Encourage multiple viewpoints and interactions with your data lineage and insights
4.  Context-aware analytics: Harness the power of graph-based AI for predictive decision support
   

1. 360 view on data lineage & insights: See the big picture of your end-to-end data-driven process

In the past few years, a substantial focus on data lineage has emerged, with clients increasingly requesting an overall view of their data lineage, questioning how it links to the reporting and risk-modelling landscape. Data-savvy clients often have access to a wide range of dashboards covering different aspects of data lineage, such as data quality, data availability or data modelling, plugged into a diverse set of data pipelines. This enables technical staff to gain valuable and granular insight into data lineage. But for senior management, who would like to monitor the end-to-end progress of data-driven reporting and risk modelling, the picture remains fractured and inconclusive.

With a non-invasive approach, and without replacing current dashboards and lineage tools, Deloitte DIM creates a data-fabric overlay. This connects existing data lineage across different sources and monitoring tools to enable end-to-end, fact-based progress tracking of data for financial and regulatory risk-reporting and modelling objectives (see Figure 1). For example, Deloitte DIM allows you to monitor the number of requirements for a specific regulatory report that has already been defined, registered, modelled, transformed and implemented by different teams.

Figure 1. The concept of the overlay dashboard

2. Effective business-in-mind implementation: Align data lineage with your business needs and policies

Data-lineage tools adopt a strong technical perspective and are excellent aids for assessing whether data-lineage projects are working efficiently towards their stated goals. However, from a senior management perspective, assessing the alignment of IT processes with strategy, mission, KPIs, policies and regulations is equally crucial. Efficiency must go hand in hand with effectiveness.

Integrating physical technical lineage with a semantic layer consisting of business requirements enables the alignment of high-level design and business requirements with actual low-level IT implementation (see Figure 2). For instance, business logic may define certain data elements as critical data elements (CDEs); technical analysis of data elements may reveal alternative technically critical elements with regards to usage, centrality and dependencies. By combining these two perspectives, you can view business-critical elements alongside technically critical elements – and therefore facilitate backward integration of technically critical issues into your business planning.

In our use cases, the semantic layer has been particularly effective in achieving ‘first-time-right’ principle and enabling responsible business-oriented progress tracking.

Figure 2. Connecting business and IT requirements for effective implementation

3. Versatility of views: Encourage multiple viewpoints and interactions with your data-lineage and insights

Data-lineage tools often focus on providing generic and rather technical views of data, using well-known visualisation methods, such as graphs or data-flow diagrams. These generic views help data analysts and scientists gain insights into technical data lineage, but for business analysts and higher-level management, they simply add complexity.

To make data-lineage insights more meaningful, we need to move towards custom views that embed the business-level logic of data-lineage clients. The tailored views should also match the background and conceptual mental models of their users to enable meaningful interactions with a complex network of data.

Deloitte DIM uses the metaphor ‘data ingredients and recipes’ to accommodate different viewpoints on data-lineage and insights. Users pick data ingredients they are interested in: for example, a specific report they would like to see the end-to-end lineage for. The selected data ingredients are stored in a user-specific data cart. Users can then choose from a list of recipes applicable to the selected data ingredients. The results of running recipes on the selected data ingredients are rendered based on a user-specific viewpoint (see Figure 3).

Figure 3. User interaction model in Deloitte Data Insights Monitor

4. Context-aware analytics: Harness the power of graph-based AI for predictive decision support

Deloitte DIM is built on top of a knowledge graph that links all semantic layers between reporting and risk-modelling requirements and physical implementation. Knowledge graphs embed a network of human knowledge and reasoning in a format understood by machines – and make it accessible to a broad spectrum of advanced analytic algorithms and graph-based machine-learning capabilities.

Knowledge graph technologies allow Deloitte DIM to leverage graph-based AI, to go above and beyond reactive insights (for example, critical element identification) and to provide predictive decision support. Deloitte DIM can proactively suggest potential new links and future patterns that business analysts may want to consider, based on the existing links and linking patterns in the data lineage.

For instance, in cases where there are items in the physical data dictionary (PDD) not linked to the logical data dictionary (LDD), different link-prediction mechanisms can be applied to identify potential links.

What we offer

Deloitte DIM is already being rolled out for several financial services clients. These clients already have data-lineage projects in place, relying on tools such as Collibra, PowerDesigner and Informatica EDC to streamline the process. Opting for Deloitte DIM was a natural decision: Not only was it an obvious fit within the existing ecosystem, it added value in seeing the big picture – connecting business requirements and technical data lineage. We accompany clients on the journey, from defining business requirements – including tailored recipes, designing initial graph models, populating graph data, writing graph queries, and developing custom-user interfaces and views – to full deployment of the solution at the end. The process starts with a proof of concept that typically takes six to eight weeks to build, depending on the complexity of the client’s underlying systems and data sources.

Please reach out to us for further information about Deloitte DIM, or if you would like to schedule a demo session.