AE Process: Agile Development Methodology
Agile Analytics Empower Clients to Make Business Decisions Without Interruption
The secret to success for Analysis Express is our Agile BI project management approach.
If we assess your needs (based on conservative estimates) and are not fully confident you will recover your investment in us during the first six months your system is up and running, we will not propose to work with you.
Business Intelligence projects are suitable for Agile development methodology. Analysis Express (AE) through its many years of implementation expertise has developed its unique version of hybrid Agile-Waterfall methodology. This takes advantage of nimble PM approach of Agile for rapid design and development type of activities that Business Intelligence projects usually go through without sacrificing the waterfall steps of implementing large scale projects. The result is lower overall risk, satisfied users and higher profitability.
The diagram summarizes AE’s hybrid Agile approach.
Event1 is the Assessment phase where AE works very closely with the client to determine high-level vision for the project and create a draft road-map of how to achieve that.
Identify Analytics Value: A Solution Inspired by the Customer’s Business Objectives
This includes, cost estimation, time, team-size, and over all effort from both parties needed to realize the vision, such as:
- Charter: Mission, Vision, Objectives, Stakeholders
- Requirements, Prioritization, GAP Study, ROI assessment
- Draft architecture, whiteboard sketch & Feasibility Study
- Planning, Funding and Approval
AE will also review any existing design recommendations or prior development efforts, that were made and will make a determination of what can be reused and repeatable.
Event 2 is related to installation and setup of necessary toolset such as Hyperion, OBIEE and ODI. At this phase, ETL connections will be tested against source databases to ensure that there are no connectivity issues. Setup of development environment will be done first so design and development can be started at its soonest.
Event 3 & 4: Design and Development, are essentially the agile development of the system which is collaborative, iterative and incremental process. AE will create and source dimensions in accumulative fashion, starting with Level 1 or Level 2 member build; and then validation by the users, before moving on the building complex dimension hierarchy and finally calculated members. Data sourcing via ETL will continue to process in parallel to multidimensional database design. At this phase, user participation will be needed the most and on a weekly basis- to validate the design.
Tasks summary includes:
- Setup environment starting with development environment first
- User interview and define requirements
- Design, Prototype and refine draft design
- Build and refine build
- User acceptance testing using limited data
- ETL effort – Recent data, Historical data
- Promote tested and verified versions to production version
- Iterative and Incremental progression towards vision
- Continue to complete setup of environments
Event 5: Rest of the events in AE methodology diagram are based on waterfall method that progresses one after another. Such as, training, testing, closure. AE preferred approach to training is to train the trainer approach. This works best for most organization since, the users have the most intimate knowledge of the data and can relate to the reports in a much more intimate way. The trainers would be trained during design and development phase so the system training can be performed by in-house user-trainers.
- Finalize release to be promoted to Production
- System Documentation
- Train end users and production staff
- System Testing
- Break-fix or final patching of any issues reported from system testing
Go-Live: One of the last step in system testing is go or no-go decision. Stakeholder participation is critical at this phase to sign off on fully tested and validated system. It is not uncommon to realize features that could be incorporated at future releases. Business intelligence projects typically evolve as the business needs shifts. Key tasks in this phase includes:
- Stakeholder Participation
- Maintenance planning, backup, restore, system monitoring process development
- Go-live decision
- Deploy to production
- Optional: Future release or release documentation
Event 6: Go-Live Support: Go-live support starts immediately after production implementation and lasts at a minimum of 1 business cycle or typically until the beginning of new period or month when, new set of data is loaded into the system. This gives an opportunity to monitor if the system is and will be performing flawlessly within production environment. Other tasks within this phase includes:
- Identify & rectify production defects
- Update documentation as needed
- Monitor system performance ie: uptime.
- Measure outcome, customer satisfaction survey
- Project closure & transition to support team.
- Optional: Cut-off of legacy system
Event 7: Continuous Improvement & Support: Long term support is all about ensuring system is performing without any issues except known issues. New issues, when identified, should be escalated and decided if a resolution can be employed. Support activities include:
- Identify & rectify new issues
- Perform enhancements
- Preform software/system upgrade
- Implement batch process to monitor system operation
AE’s agile approach to development methodology allows for minimal disruption to the organization while development of the new system takes place. AE’s process delivers results by providing rapid and continuous development of software and systems with cooperation between stakeholders and users. This approach easily adapts to changing requirements throughout the process by measuring and evaluating the status of a project. The measuring and evaluating allows accurate and early visibility into the progress of each project. This process ensures users are happy with the results, and the AE team satisfies the client’s needs.