The world of software estimation is struggling to keep pace with the changes in the speed of software delivery. More than 80% of major federal IT projects now using Agile processes, planners are under increasing pressure to keep up and add value. Providing accurate cost, schedule, and labor estimates is nearly impossible without a full understanding of the technology solution to be delivered and tools that reduce analysis cycle time. Federal IT Project Managers are simply not interested in funding 6-month Business Case Analyses, and often pose different questions to estimators than simply the long-range budget forecast.

A key aspect of this change involves functional size measurement. The industry has moved away from source lines of code analysis, and Agile projects speak in terms of features, epics, and user stories. Function points continue to be well-suited to Agile estimation due to being standardized and following a well-established set of rules that can be readily applied. However, functional sizing is a tedious and rigorous process. Fortunately, advances in Artificial Intelligence and Natural Language Processing can be utilized to cycle through hundreds of requirements or user stories and translate planned capabilities into software size. The analyst can then focus on adjusting the measurement and requirements elaborating.

By taking advantage of advances in computer science, we as estimators can continue to add value to projects by providing rapid estimates, identifying impacts of feature changes, responding to Project Manager requests, and advising our clients on technology alternatives.