The incredible combination of agility and Artificial Intelligence

26/03/2021 185

Agility as a business attribute arose from the need to operate in a predictable manner even in the face of extreme complexity. In particular, in companies of software development, agile methods have been promoted as opposed to more rigid cascade development models. The first interactive and incremental methods of software development date back to the year 1957. Since then agile methodologies such as Scrum, XP or Kanban have been developed. In 2001, participants in the development of agility met in Snowbird, Utah, adopting the name of Agile Methods. Subsequently, some participants formed the renowned Agile Alliance in 2016 as Agile Glossary.

Hirotaka Takeuchi, dean of Hitotsubashi University, and Ikujiro Nonaka, a professor at the same university, introduced the Scrum concept in their article “The New New Product Development Game,” published in Harvard Business Review in 1986. Scrum is a framework for managing agile software development. Its name comes from the advance in Scrum formation of rugby players. Scrum is designed for teams of three to nine developers who divide their work into two-week cycles (sprints), monitor progress daily in fifteen-minute meetings and deliver usable software at the end of each sprint. The main roles in Scrum are: product owner, Scrum master and Scrum team.

The origin of Kanban goes back to the 1940s, when Toyota developed new control systems to achieve waste-free production focused on delivering maximum value to customers. Kanban has adapted to knowledge management as a visual process system that aims to manage the work, balancing the demands of customers with the available capacity to improve the management of bottlenecks. It consists of five elements: (1) visualize, (2) limit the work in progress, (3) direct and manage the flow, (4) explicit process policies and (5) use models to capture opportunities for improvement.

In coming years, agility methodologies will be integrated with Artificial Intelligence and Machine Learning projects, generating more advanced capabilities.

Software engineer Kent Beck, in his book Extreme programming explained: embrace change, published in 2000, introducing the software development extreme programming methodology. This concept, also known as XP, is an agile process whose objective is to improve the quality of software and the ability to respond to the changing needs of customers. The origins of XP go back to different ideas that appeared since the early 1960s in projects such as NASA’s Mercury. Its current use facilitates frequent launching of developments in short development cycles, with points of revision and control connected to the needs of customers.

In 2018 Ayman Sayed published an article in Harvard Business Review entitled “Using AI and Machine Learning for agile development and portfolio management” explaining that the business world was beginning to explore the possible application of Artificial Intelligence and Machine Learning for agile development, testing and even portfolio management. These practices could answer questions like, for example, if you did not have to rely on people or traditional software systems how much faster could you get to market. Product or software delivery schedules could also be predicted much more accurately.

The article proposed the integration of intelligent machines to look for patterns, code anomalies, changes in equipment production or to analyse delivery plans. Companies should create a way to dynamically visualize, plan and track work across their entire portfolio. Delivering products to market with the speed and accuracy that customers demand means that everyone in the business needs access to the right data, at the right time, and in a format that insights are accessible and actionable.

In a recent article by Kathleen Walch for Forbes entitled “Why Agile Methodologies Miss The Mark For AI & ML Projects” the author explained that progress must be made in order to connect agile methodologies with Artificial Intelligence and Machine Learning projects. The article mentions methodologies such as CRISP-DM (Cross Industry Standard Process for Data Mining) consisting of an open standard model of the process that describes the common approaches used by experts in data mining. Another methodology mentioned is CPMAI (Cognitive Project Management for AI).

In conclusion, in the coming years agility methodologies will be integrated with projects with new exponential technologies to implement more advanced capabilities such as self-service business intelligence or predictive analytics based on Machine Learning, Artificial Intelligence and automation.