Seventy million documents were an asset. The troubled project needed another view.
A financial-information undertaking had run for several years and was millions over budget. Its pay-per-use information system involved more than seventy million documents and a mainframe migration.
Howard Pearlmutter was asked to examine alternatives, including cancellation, disposition, or resuscitation.
His investigation connected technical architecture, quantitative analysis, parallel search throughput, markets, and the perspectives of stakeholders and teams.
A troubled implementation did not answer the value question.
The existing information asset and the current route toward it deserved separate examination. A decision about the project could otherwise become a decision about the asset by implication.
Howard's work produced a new direction that preserved value and opened alternatives. Investigation, recommendation, and rearchitecture planning took one week. A successful launch followed later; it was not a one-week launch.
Throughput needed a purpose.
Searching a large corpus is useful because it helps people obtain the information the service is intended to provide. Architecture, economics, and market purpose therefore belonged in the same examination.
That relationship connects this case with cachemetrics.com's broader question: what useful work does a favorable technical measure represent?
A system can be busy, fast, or highly reusable while the important uncertainty remains elsewhere. A baseline and a clear finish help keep the intended result in view.
Experience worth bringing to a difficult comparison
Howard's development and consulting clients include NASA, Apple, HP, CSC, Capital One, and the Financial Times. His practice combines detailed technical work with the larger undertaking that gives it value.
Explore architectural comparison, systems review, or Howard's background.
Before treating an investment as one indivisible success or failure, examine which parts of its value and architecture can still be put to work.
