A large client undergoing a reorganization needed visibility on its business practices, products, services, staffing levels, and workloads. To communicate and plan, stakeholders across the organization required data-driven insights about current and future operational needs.

Our Challenge:

• The organization’s internal data capture was implemented in varying and often inconsistent ways among the organization’s operating units.
• Data definitions reflected complex internal guidance and rules. As a result, data meanings were not readily interpretable or explainable in plain language.
• Status quo data insights were being produced in an environment relying heavily on opaque and complex handling methods. This led to time consuming data extraction and results reporting.
• Data quality controls were needed to identify problem areas and verify the strengths and weaknesses in the client’s data methods.
• Although large amounts of data were generated at operational levels, legacy databases and over-reliance on basic tools (spreadsheets) hampered the targeted insights that decision makers needed most.

Our Solution:

The solution was layered into several steps:

1. Define the business scope

• Identify existing operational issues and reorganization questions
• Define business processes
• Define how the organization leverages IT systems and contribute data to the business

2. Identify and assess available data

• Gather and generate data generation across any/all relevant sources
• Align data to business processes and requirements

3. Refine the data

• Cleanse and validate the large and disparate data sets to a level capable of addressing the critical issues and questions in staffing, funding, and workload

4. Apply a tailored machine-learning approach (k-clustering vector quantization) to identify actionable insights

• Posit overarching hypothesis regarding business performance and management
• Identify operational issues and verify conclusions leveraging the data

5. Integrate data insights and empirical evidence to update processes and develop a way ahead

• Develop data-supported actions to address process deficiencies
• Undertake data stewardship to monitor and manage staffing and funding process updates

A visual summary of our solution is provided in Figure 1.

Figure 1. Data Solution Cycle