Predictive Project Analytics Predictive Analytics AI for Project Management

AI for Project Management: key innovations having an impact

In his latest blog, Cora CTO Pat Henry takes a look at AI for Project Management and how five key innovations are already starting to make a difference when it comes to managing project portfolios.

Artificial intelligence (AI) has opened up a brave new world when it comes to managing project portfolios. A lot of the new advances are grounded in common sense. Here we take a look at how five key innovations are already starting to make a difference.

1. Aim

Leverage the vast amount of data available within your project portfolio management (PPM) platform to identify what attributes are important in predicting the success of a project. What is it that great project managers do? Is there a way you could use the data within your PPM platform to predict the success of a project or to prompt the user to perform some actions which could potentially turn the project’s RAG status from red or amber to green? Or are there any indicators that identify that the project is going to fail, and we should cut our losses and move on?

2. Motivator

Currently, in your Cora PPM platform, a lot of operational level data is gathered – from various logs, registers, workflows, uploaded documents, audit trails, etc., but this is not analyzed holistically on a project or portfolio level. We rely on the project manager to interpret what all of the data means for their project, and based upon their insight, we update the project RAG status, update various registers and logs and report upwards if there are any issues, etc.

Building upon our in-house expertise in the area of AI and machine learning (ML), we saw a gap in the market for a system that could automate some of this manual analysis. By having insights on the project and portfolio status automatically generated, we knew we would be empowering the project manager on the ground to make necessary decisions faster. A knock-on benefit is that this frees them up to spend their time on value-add activities rather than admin tasks.

3. Proof of concept

Cora Systems powers best practice enterprise PPM. One area in which we are currently growing our offering is around financial information, cost books, financial forecasting, etc. Forecasting lends itself to AI and ML techniques. Specifically, we focused on the accuracy of the EAC (estimate at completion) forecast. The reason being is that forecasting cost accurately is important to drive precise revenue estimates and hence the EAC shouldn’t change over the course of a project.

To get an understanding of EAC deviation, we needed to identify the data points (levers) which impact EAC adjustments. This information can have a bi-fold impact: first, it can be used by AI and ML techniques to give percentage confidence about the predicted EAC deviation/adjustment, and also it can be used to promote process changes in our client companies by informing them of the attributes which are more important and need to be tracked/monitored.

Over the course of four months, we used that problem to upskill our technical team on the intricacies of data extraction, cleaning, and the basics of AI and ML. The team got exposure to supervised and unsupervised learning algorithms as well as reinforcement learning techniques and learned the types of problems and data which were best suited to them.

4. Natural language processing

Another area of research and innovation within Cora’s platform which we are actively involved in is around natural language processing. Over €20bn worth of projects, in over 50 countries across the globe are managed using Cora PPM daily. At any one point in time, there are over 300,000 live projects operating in Cora PPM.

During the operation of these projects, a significant amount of free text is entered into the system, along with a large volume of documents, some of which also include handwritten notes. Within Cora PPM at the moment, all of this information is stored in its raw state. Project managers must log in to the system and manually select to view this information before making their decision or updating the project status. The natural language processing service will parse, extract and annotate all of this raw data, enabling machine interpretation.

Having the ability to run sentiment analysis (looking at text someone enters and trying to gauge mood based on what is inputted) on a project or a portfolio will enable our clients to ensure that they are focusing their efforts on the correct projects and portfolios. The system will be able to raise early warnings where the perceived sentiment does not correlate to the reported project or/and portfolio status, and potentially escalate risks or issues.

5. Strategic portfolio management (SPM)

Currently, we are building out our new strategic portfolio management product Cora SPM, which will build upon the strong foundations of Cora PPM, and we will apply advanced data analytics to convert operational-level data from project and portfolio management platforms (such as Cora PPM) into insights.
Cora SPM will have the capability to predict the success of projects, identify which projects should be included in a portfolio, and, ultimately, guide project and portfolio managers to ensure that they achieve repeated success.

Incorporating advanced data analytics and machine learning into the project and portfolio management space, when combined with our in-house domain expertise, will give Cora a significant competitive advantage, especially in our strategy realization offering.

Find out more

If this area is new to you why not scroll down for more resources on AI for Project Management.