Architecting and Data Modeling
Data Warehouse Architecting is based on the business processes of an organization. We take into consideration data consolidation across the organization with adequate security, data modeling and organization, query requirements, meta data management and application, and planning a warehouse for optimum bandwidth utilization and full technology implementation. Our Data Warehouse Architecting is a multi-faceted process involving:

• Process Architecture: defines the number of stages and how data is going to be processed to convert it from raw transactional data to information which can be consumed for decision making by the end users
• Data Modeling: is perhaps the most important stage of architecting a data warehouse as the ability to integrate and deliver information across the enterprise is largely dependent on the quality of the data model architecture below. We define and analyze data requirements needed to support the business processes of an organization, record them as a conceptual data model with associated data definitions i.e. not just defining the data elements but the structures and relationships between them. We see Data Models as progressive, in other words as a living document that changes according to changing business, and recommend that the model be stored in a repository, so that it can be retrieved, expanded and edited over time.
• Technology Architecture: covers the analysis we provide to the scalability and flexibility of the BI Solution. This largely depends on the organization size, nature of business, business requirements etc. Also, we derive technical architecture from process architecture, metadata management, and other requirements like business rules, security considerations, tool specific needs etc. In this step we also look into carious technology standards like Database management, connectivity protocols, middleware, network protocols and related technologies.
• Information Architecture: is where we look into the process of how to translate information from one form to another in a step by step sequence. This would enable us to map how to manage the storage, retrieval, modification, deletion, etc. of data in the data warehouse.
• Performance Optimization: Many resources, we believe, come from the software architecture which help in determining performance. The other consideration we have regarding performance is the workload. Simply put, if we design a solution that has enough resources to complete the workload on time, then the performance will be optimal.

Data Integration
Data Integration involves combining data residing in various sources and providing end users with a unified view of enterprise data or as it more commonly called, a “Single Source of the Truth (SSOT)”. Data Integration has been increasing importance as we see data volume explosions all around us today. Also, the increased adoption of Transaction Processing systems like ERP, SCM, CRM, etc. means that data is distributed between multiple data silos and getting an enterprise view of data becomes a time consuming and complex process. This process is doubly cumbersome as transaction systems need to be online 24x7 and running reporting queries on the transition databases increases the latency for the report users, as well as the transaction system users.

The solution, therefore, is to create an SSOT or reporting database which has (a) all of the data elements required for getting enterprise information nin one silo and is (b) separate and unconnected to, transaction databases. Based on the purpose and scope, this reporting database could be called an Operational Data Store, Data Mart, Data Warehouse etc.

GrayMatter has a lot of experience in delivering complex Data Integration solutions to our customers. Our services in this area provide for the Extraction of data from the source system, running complex transformation logic on the data for standardization, aggregations etc., and finally loading this modified data into the reporting database. This process is typically ruled by the Data Model created while architecting the solution.

Data Governance
Data Governance is the meeting of Data Quality, Data Management, Data Policies, Business Process Management, and Risk Management surrounding how data is handled within an organization. It therefore a complex program that requires a fine balance of using soft skills of managing people, committees, management, workforces; and the strong analytical skills to “get your hands dirty’ with your data model, data process, metadata etc.

GrayMatter as part of our overall service offering is experienced in creating and managing Governance programs at some of our clients. Typically (and simply put), our Data Governance process scope is focused on creating a data governance maturity model, which could be and most often is, different for different customers based on various factors like business process rules, industry, size etc. The data governance model typically looks at data elements within the organization and deciding (a) if it needs governance and (b) if yes, then how would the process of moving this data element from ‘no governance’ to ‘fully governed’ be defined.

Our framework for data governance