-
Renal Information Exchange Group
Situation
RIXG is a steering group that has a remit to improve the flow of information for renal (kidney) patients.
Problems
RIXG wanted to implement a web site that would collate patient data concerning medical tests, drugs and other data from renal units around the country. Patients would log in to see their results, increasing patient involvement in their condition management and placing a lower burden on doctors to report results directly to patients.
Solution
Worth Solutions worked with medical experts from RIXG as well as local medical and administrative staff from the renal units to define the processes needed to extract data from the systems to a central database and web site. The customer as a renal patient was defined. What was of value to a patient was clarified. With this information a value stream for the transfer of data and its display was formulated. The stream was optimized to maximise the flow and minimise the waste in all steps. Quality of data was a big concern, so techniques to build quality into the data processing without reliance on inspection were very important.
Results
The process analysis of customer value, flow and data integrity meant that when the implementation of the technology came along it was a smooth as possible. The time for implementation was vastly reduced due to a clear understanding of how to maintain data quality and what was of most important to the renal patients who would use the web site.
-
Serco
Situation
Serco is a global outsource services company. It was recognised that in the Defence and Aerospace division much work needed to be done to develop a repeatable method of improving service delivery for any Serco client contract.
Problems
Processes were not understood, let alone mapped in Serco client deployments. Virtually all Serco processes were fragmented. Work did not join up at a service, technology or data level. This hampered good practice, and stopped Serco delivering the excellent service that is their ethos.
Solution
Worth Solutions joined with Serco to staff a project team to develop a repeatable method to understand, map and improve Serco client service delivery processes. A methodology was put together to map service level and system level processes. Coupled with Worth Solutions’ extensive process experience we brought in-depth understanding of technology solutions and implementation.
Results
The output of the project was a method that Serco can apply to any client contract to understand, improve processes. This will lead to extensive financial savings, higher client satisfaction and an increase in winning tender bids with a more mature method for running services. The methodology is currently being applied a Serco client.
-
JP Morgan
Situation
The group built software for the internal Equity Derivatives customers, groups of traders and their eternal clients. The applications were web based utilising a Java/J2EE platform.
Problems
Code quality was low, many releases contained bugs that would cause releases to be pulled just before implementation or immediately after when problems were discovered by clients.
Solution
Concentrating on the principle of Customer Value, we changed the requirements gathering to be just-in-time so new features and bug fixes were worked on in two-week slots. Future features were ignored and code was written only to implement the task at hand. A test harness was implemented such that code was written test-first. No test, no code. A standard format for code was agreed and an automated way of applying it made it easy to use for developers.
Results
The formatting standard and test first practice increased the quality of code. Far fewer bugs were made into the system. The two-week iterations meant that smaller changes were released but more frequently, this increased flow of features to clients. Coding features just-in-time and ignoring work far into the future meant that the system was much simpler and coupled with reworking code mercilessly this meant that it stayed simple, thus drastically reducing the cost of each change. Bottom line – releases went out more frequently so clients saw value earlier and the incidence of pulled releases went to almost zero in 7 weeks.
