We deliver BI with Agility
Our AgileBI approach allows your team to deliver while embracing frequent change and not freaking out.
There are five distinct capabilities we deliver in an AgileBI way.
Discover true requirements
Our AgileBI approach combines a number of methods to help close the gap between business user requirements and data warehouse design.
Via a quick visioning process, we confirm that all stakeholders have the same expectation of what will be delivered. This includes what is expected to be in and out of scope. Often this results in us reconfirming who the stakeholders for the deliverables are
The Business Event, Analysis and Modelling (BEAM) method is used to elicit business processes and the underlying data that supports these processes. Then we identify the data that is required to meet the modelling and visualisation requirements.
Using modern data discovery tools we profile and explore the data to confirm we have access to the right data to deliver the requirements.
Using business rule notation, we capture the intricate combinations of rules that are required to cleanse data or calculate new inferred data attributes.
We document the metric formulas, so we can confirm we can build the required measures and key performance indicators with the data identified through the BEAM process. If we find something missing we refine the data requirements.
We use wireframing techniques from the world of application development to build storyboards, confirming the visualisation and interactivity requirements, before we start building reports, dashboards or information products
Once all the various requirement types have been gathered, we create the backlog of stories that identify the things that need to be delivered by each of the AgileBI project sprints.
Attend our ‘An Optimal Introduction to BEAM’ 1-day workshop to help you get started in using BEAM to gather data requirements from your business users.
Vision and Scope to agree expectations
We use a rapid visioning and scope process, which takes less than an hour, to set and agree the expectations and what will be delivered next.
We use the vision story to confirm the intent of the Information Product, creating a simple and focussed statement of expectations
We identify the items that are expected to be in and out of scope. These enable us to identify when a major change of scope is being requested during the delivery cycle
6 X 6
We use a technique comprised of six sentences of six words, quickly identifing the key principles, constraints, dependencies and assumptions that drive the expectations
We use the stakeholder onion to identify the key stakeholders, who have input into the requirements for the Information Products and identify them as Customers, Influencers, Reporting Owners or Rule Makers
We use the Context map to represent visually the flow of data from the identified source systems to the identified downstream information systems and users
We automate data integration to enable rapid change
Our combination of Data Vault modelling, and our open source data warehouse automation engine, simplifies the data flow and transformation process, allowing data models to change and be extended with agility in the future.
We architect distinct data layers to ensure it is clear where data is augmented, and we utilise data automation techniques, to ensure we can reconcile data at each layer and change business rules with agility
CHANGE DATA CAPTURE
We use change data capture (CDC) capabilities to automate the acquisition of all changes in source systems with minimal effort and to ensure we never lose any detail
TIME VARIANT LAYER
We store all changes, for all data, for all time to ensure historical data can be accessed at any time for analytical, behavioural modelling, and to enable business context and business rules to be applied iteratively
DATA WITH SLA
We identify data that has an expected SLA for delivery of changes to ensure we automate it, versus data with no SLA that can be refreshed at any time and potentially by anybody
We apply the Data Vault modeling technique to ensure the data models, and business rules can change and be extended with agility in the future
VIEWS, VIEWS, VIEWS
We use database views to virtualise data structures, enabling rapid prototyping, management of rapid change, and providing users with early access to data for validation
We use a combination of an associative in-memory engine combined with dimensional, denormalised and summary views to represent the data in the most effective way for the users and tools accessing it