Effort estimation remains a relevant challenge in software engineering, even after the consolidation of agile methods, DevOps, continuous integration, continuous delivery, and incremental delivery practices. In 2005, we presented an approach based on the Experience Factory concept to improve software effort estimates through project characterization, historical productivity data, and organizational learning supported by metrics. Two decades later, the central question is no longer whether that approach can still be applied, but how its principles were preserved, transformed, or absorbed by contemporary software engineering practices.
This talk revisits that original proposal from both practical and critical perspectives, connecting Experience Factory, Goal-Question-Metric, historical productivity repositories, functional size measurement, and continuous improvement with modern agile environments supported by product backlogs, source code repositories, continuous integration and continuous delivery pipelines, retrospectives, DevOps metrics, engineering analytics, and data collected from the daily flow of software development and operation.
The central argument is that the Experience Factory did not become obsolete. Instead, many of its core principles were incorporated into modern software engineering practices through continuous feedback mechanisms and data-driven engineering approaches. At the same time, several challenges identified in the original work remain unsolved, including contextual characterization of projects, degradation of organizational memory, and the risk of accumulating metrics without generating actual organizational learning.
The presentation discusses lessons learned from the original study, practical limitations found during industrial adoption, and possible reinterpretations of the Experience Factory concept in contemporary software development ecosystems.
Date and Time:
June 29th, 2026: 10 a.m. Eastern Time (Time Conversion)
Presenter:
Ricardo Ajax Dias Kosloski is a Professor at the University of Brasília, Brazil, working in Software Engineering, Software Measurement, Software Quality, Process Improvement, and Software Effort Estimation. He holds a Master’s degree in Knowledge Management and Information Technology from the Catholic University of Brasília, where he developed research on continuous improvement of software effort estimation using the Experience Factory concept.
He is a Certified Function Point Specialist and has professional experience in software factories, software process improvement initiatives, Capability Maturity Model Integration level 3 and level 5 implementations, Melhoria de Processo do Software Brasileiro initiatives, software metrics programs, and organizational productivity analysis. He has also worked with agile software metrics, software testing metrics, and measurement models for outsourced software services.
His doctoral research at the University of Brasília investigated the use of digital government services by older adults, with emphasis on digital citizenship, software product quality, accessibility, trust, security, and the barriers that affect the use of public digital services by this population. His current and future research interests include software engineering, digital government, organizational learning, software quality, and the use of digital government services and software systems by older adults.