Machine Learning to Predict Project Trends — WeTrack Trend Analysis

Our CTO, Tom Mason, Talks AI

Our CTO, Tom Mason, discusses the many uses of Artificial Intelligence and its variants, including Machine Learning, and how these are being applied at WeTrack to help our clients manage their projects better.

Machine Learning to Predict Project Trends

Tom was the architect behind our machine learning framework which allows users to see and analyse trends in their project and risk management which they might otherwise miss.

Let’s start with the basics… what is machine learning?

Machine learning is a backbone of one of the biggest topics in computing at the moment. Most AI systems will be driven by systems that can be taught in some way to “learn” to perform in increasingly optimal ways. That is, they can be trained. At WeTrack we started by giving people the tools to manage their projects and the challenges surrounding them. One of our current goals is to give people the ability to manage their challenges pre-emptively. We want to help them identify problems before they happen, using trend analysis driven by machine learning.

Why did you think it would be useful to apply ML to WeTrack?

Fundamentally, projects in many different industries share similar challenges. Risks to completing a project on time can come from different sources but if we know how to codify and detect these before they occur, we can teach a system to predict when problems may be about to occur. We noticed great commonality between clients we’ve worked with over the years with WeTrack and this led us to look at how we can improve the system using the lessons learned from others.

How does the trend analysis work and what exactly is the machine learning in this instance?

A simple example would be the correlation between risks in one category of the project and potential delays in another. In all project management systems, the ultimate quality of the data depends on the users who codify the connections between the parts of the system. If the catering team has delays in completing their project components, could the quality assurance team be unable to complete their overall work? WeTrack aims to remove the need for users to code these connections themselves but suggest how patterns may emerge based on the global knowledge of the system as a project management tool. We apply machine learning in this aspect by improving the recognition of these connections each time the tool is used, on a global scale.

Do you see any further applications for AI or ML in WeTrack in the future?

Ultimately WeTrack is designed to ensure our clients deliver their projects on time, efficiently and smoothly. Our trend analysis tools are in their infancy but with the nature of the projects we work on, it is only a matter of time before we look at introducing the support of machine learning algorithms into all parts of the platform.

Finally — what excites you and what worries you most about artificial intelligence?

A truly intelligent tool could evolve beyond the realms and rules governed by its creator — a concept marked by many as a reason for concern with AI. But ultimately we stand to gain from using basic machine learning to assist in the recognition of otherwise hidden patterns in the challenges facing the project management world. Before we take on SkyNet — we aim to help you deliver first!

Originally published at www.wetrack.com on January 15, 2018.