About The Author
Arman Kamran 12 articles
Residence: CA North York, Ontario
Enterprise Agile Transformation Coach, CIO and Chief Data Scientist

more about Arman

All Authors


The Unified Project Management Dictionary

more terms

AI and Project Management

Artificial Intelligence (AI) has always been a fascinating subject to humans. Picturing a machine capable of thinking independently and making decisions without our intervention has traditionally been so attractive to us that many books and plays have been written about that and many movies and cartoons have been made since the dawn of moving pictures to this day.

On a personal experience, after a long period of Sci-Fi-based AI enthusiasm during my school years, I finally got introduced to AI during my undergrad years while studying at a sister university of MIT, working on a joint project with MIT (as my graduation project) which would use AI in navigation decision making based on learning the traveled path in order to find its way back based on that learning alone (without using GPS).

Back then, we did not enjoy today’s abundance of cheap yet super strong processing power and training data, so most of what we would do was simulated in labs with painstakingly slow CPUs and mathematical modelling we had to cover partly computer assisted and partly manual.

My connection with AI, despite its initial super-boring and painfully slow status, continued through my Post grad education and as processors became stronger and cheaper and data became more available, it became more practical and usable in day-to-day life in a variety of market sectors.

Today, I serve as a Chief AI Strategist of a Machine Learning (ML) Software boutique with bases of operations in 3 continents.

Your new AI-based Project Assistant

AI/ML is designed to enable independent thinking outside human mind. This means the cocktail of algorithms used in an AI/ML solution is designed to learn from the training data provided to it, so it can repeat the thinking practice on its own without human intervention.

This data can be provided directly by a file filled with training values, or by having means of watching you during an activity and learning from what you do and how you are doing it.

It can also expand into watching your mistakes, other people’s mistakes, errors in information and all sorts of problems that you may have to deal with, so it can try to avoid them through a variety of ways (finding better practices, double checking the data, sending reminders to participants to ensure they provide their reports on-time with the needed accuracy and more).

As it would be expected, AI/ML usually starts with taking care of repetitive easier tasks – technically called as RPA (Robotic Process Automation) - that would need to be done but would take away your time doing it despite their lack of complexity. 

Studies show that more than 50% of a PM’s time is taken away by Administrative tasks that can be delivered through an AI/ML solution.

At this point AI/ML acts as your assistant-in-training.

As your AI/ML solution learns through assisting you with the easier tasks, its level of sophistication grows up and now can extend into some more complex parts of your work.

This would allow the AI/ML solution to not only save your time from being spent on less valuable tasks, but can also serve in predicting the trends that are forming up or are hidden in the flow of your project and raise awareness on the risks hidden in the works.

AI/ML can even go further into studying how the teams working in your project are functioning, what is their level of efficiency, what are the parameters impacting that and what is the health level of your team structure.

It can even go deeper to the level of watching how each individual is performing in the team, what is the trend of their efficiency and accuracy and even assist them with needed reminders, knowledge transfers and assisting them do a better job.

Here are some of the areas that AI/ML solutions are already actively serving projects and expanding into further capabilities:

  1. Data visualization and Decision Support insights helping the PM and the Steering Committee make better informed decisions and track their outcomes.
  2. Data enrichment (through addition of useful Meta-Data to provide more insights on the otherwise-raw-data acquired through your feeds and stored in your repositories).
  3. Administrative services (from scheduling meeting needed to sending reminders and follow ups and collecting status data across the teams and individuals).
  4. Budget and Schedule tracking, trend observation and prediction of outcomes and raising alerts when needed.
  5. Teams and individual performance tracking, monitoring, trend analysis and improvement planning. This also can expand on detecting team morale levels and suggesting workload balancing, training sessions and adding team building activities and morale boosting technics.
  6. Providing Bias-Free, high quality analysis on the current and trending project status.
  7. Task / scope item re-assignments, Sprint revisions and live load balancing on the teams based on best performance directives and shifting live business priorities in re-action to market changes or strategic updates.
  8. Integration into the delivery pipeline and observing the delivery eco-system from concept to product to find bottlenecks and identify synergies and to provide recommendation for improvements.
  9. Performing better Risk and ROI and Value analysis using the enriched data and live data collection across the project teams.
  10. Integrating to Project Resource Planning and even the HR system to predict the technical specialization needed at the start and throughout the life of the project.

Challenges on the road

Those of you with experience in project management are fully aware that there has never been any “nicely done” large project in history. Every one of them goes through numerous pitfalls, challenges, conflicting priorities, missed milestones and end up finishing with a variety of delivery accuracies and completeness.

All of theses projects have one common issue:

Data collection and reporting is always messy and inaccurate and delayed, leading projects not having clean, complete and adequate amount of data to properly train and feed their prospective AI/ML solution.

This is the main challenge in training your AI/ML solution into a functioning state to kick start assisting you and your team.

Since this is an on-going issue through the life of a project, providing the AI/ML solution with the needed clean, correct and complete live data feeds that would allow it to make sound decisions and achieve high quality predictive analysis of the trends, is another big challenge.

Solution providers have started integrating the AI/ML solution into the project management tools, making it responsible for collection of its needed data, polishing and enriching it and using it to feed its decision making and predictive abilities.

This would bring visible improvement in efficiency and accuracy of these solutions, and would cover for many error-prone data collection practices where humans traditionally were introducing mistakes in entering the data or misinterpreting the information they had while translating it into the feeds used by the system.

A voice of concern

Now that I have listed the array of benefits a project, its PM, committees and team members can receive from incorporating AI/ML solutions in their day to day activities, I should also share with you a caveat!

Over the past decade, I have noticed, raised and shared my concerns over the serious risks that the currently uncontrolled and unlimited experimentation with AI as a booster to productivity and service provisioning is building up at a global level.

Artificial Intelligence and its child, Machine Learning are designed to remove dependency on humans’ brains as the source of thoughts and decision making. ML is developing extremely fast in every area that formerly was dominated by humans and has proven its ability to defeat humans in every sector it has penetrated.

We are not physically strong species. Without tools there is very little we can achieve in the physical sense, but up to this era, we were always the most intelligent specie in the situation, and as a result we always managed to pull ourselves out of every mess we brought upon ourselves.

With AI/ML, for the first time over the past thousands of years, we are no longer going to be the smartest element at the table and this puts us in a very disadvantaged position.

I am not trying to raise an alarm on the Sci-Fi side of AI, fearing the raise of the machines and threating our survival as a specie but that of what we are blindfoldedly bringing upon ourselves at a global and local level.

At its current state, AI/ML have become the main topic of a “Gold Rush” movement at global levels and businesses are pushing towards “Total Automation” through using it, removing any dependencies on human brain.

Let’s not forget that at the end of the day, no enterprise has any kind of commitment to their employees prolonged employment. They are and have ever been, only accountable to their shareholders and have always seen human resources as a const factor and layoffs are usually their first reaction to any financial hardship that gets serious.

In this case, since their competition in the market is going to lower its costs significantly through automation and lowering the payroll costs, that is going to be the path they will have to take if they want to stay in the game.

As much as industry leaders have been trying to brush off these concerns as being “far-fetched” and “too-distant-into-the-future”, over the past couple of years we have witnessed an accelerated pace in AI/ML automations taking away job growth in many sectors by simply not hiring human in many areas, or starting to shut-down human-driven divisions and replacing that with an entirely AI/ML driven model.

We have witnessed how Call Centers, which traditionally employed hundreds of humans each are now successfully reducing their staff members by over 75%, giving the repetitive cases to their AI/ML solutions to cover.

We are now seeing enterprises, quietly shutting down operations, laying off thousands of people only to cover the services through automated solutions.

We are seeing giant online sellers building order fulfillment centers capable of covering hundreds of thousands of daily orders, hiring only a handful of robotic and AI/ML experts to watch over and enhance the entire operation, while those centers used to have thousands of employees up until a couple of years ago.

This is very different from the time that automation in manufacturing took away millions of jobs yet left the hope of training the humans into more mind-intensive work elsewhere.

AI/ML is replacing the dependency on human brain, and there is no training in the world that can enable humans to somehow complete with these solutions.

Base on a calculation we did merely 2 years ago, AI/ML was able to take away over 1 billion jobs. Now imagine with the accelerated advancement in this sector ever since, that number has now raised to much more than that.

This un-checked, under-regulated, open-season approach to pushing for “Total Automation” in a world where we are at the verge of population explosion will only exacerbate the un-employment gap that is widening fast and raise the percentage of population under the poverty line.

Unless, at a global level, all governments start establishing the missing guidelines and regulations, this runs a very high risk of creating a global imbalance that can someday pass the point-of-no-return and push our global economy into an unprecedented chaos, leading to threat of world-wide clashes that can happen as early as next two decades.


Published at pmmagazine.net with the consent of Arman Kamran