Gil Broza’s scientific contributions

Extreme programming (XP) literature and discussions often view successful projects only as customer-driven product development: planning, coding and testing an unfolding series of prioritized units of vertical functionality. I claim, however, that a successful project also requires a prospering community, comprising an introspective group of committed professionals communicating effectively, and using a well-understood, stable process. Weakness on any of these fronts presents a high risk of failure; therefore, I advise every XP project's members to actively engage in building their community, such that it reaches its critical level of development already by the first internal release. To help in this endeavor, I provide a comprehensive list of activities and attitudes to practice and avoid during the first release.

Affinium Pharmaceuticals engages in early-stage pharmaceutical R&D and molecular biology production processes for internal and external programs. This business requires significant informatics support in terms of small- and large-scale software, tool integration and data management. Obtaining suitable software is difficult due to customer diversity, rapidly evolving unique needs, vendor offering and high costs. Adapting the XP approach and practices for this situation, Affinium’s Informatics group has successfully developed in-house software that has kept up with the science. I describe notable accomplishments, and lessons learned along the way. I propose that a small in-house group of domain-aware developers, using a customized version of XP, would achieve better results than external providers, despite limited access to resources. In closing, I suggest that this structure and methodology are generally applicable to dynamic research, development and production environments.