The Most Complex Business Management Task Ever

The conventional wisdom – to start small, pick the low-hanging fruit and then scale up – has given way to an acquiescence that organisations need a fundamental shift in their approach to data to do even that well. Moreover, at times, starting small makes scaling up significantly costlier and more difficult. This makes scaling-up without a prior holistic AI plan either less probable or altogether impossible. Even though AI onboarding has become a critical task, most organisations are still not sure where to start, and seeking professional help might make a huge difference.

With the accelerated automation brought about by the Covid-19 pandemic, the big question on everyone’s mind remains when AI onboarding will eventually mean that the bots are going to take all human jobs and how this will happen. While most experts agree that for some considerable time to come there will always be roles for humans, even in a machine-powered future until AI has managed to surpass all human capabilities, the immediate questions remain what those jobs will be and what they will look like, what skills they will require, how companies can best prepare their workforce to acquire the skills to do such jobs, and how humans and AI will eventually work together. This begs the even more fundamental question of whether humans are managing the machines, or whether the machines are going to be managing the humans. The answer to this question would usually dictate the managerial approach to AI onboarding, as well as the technological and leadership pathways adopted by the organisation, and will eventually have regulatory, management and process-related implications on the firm’s AI growth pathway and the firm’s growth more generally. It stands to reason that the greater the degree of organisational focus on people helping AI, and AI helping people, the greater the value that can be unlocked by the AI transformation for the foreseeable future.

As technology improves, robotic automation projects are likely to lead to some job losses in the very near future, particularly in the offshore business-process outsourcing industry. As a general rule, if you can outsource a task, you can probably automate it with relative ease.

While AI is in itself a complex and costly management task to implement, once implemented it can increase the value and efficiency of management. Indeed, there is no aspect of management, from planning, decision-making, control, execution and monitoring that AI has not impacted, oftentimes drastically, for those organisations who have decided to invest in it.

When it comes to planning, AI is being used to create predictive analytics and to provide real-time alerts. It can help identify emerging trends, manage and optimise stock-keeping units, assist with supply chain planning, as well as plant and facilities maintenance management. It can also optimise resource allocation plans and help with project management throughout the entire project management lifecycle by ensuring realistic resource allocations commensurate with the tasks and by identifying tasks that are likely to slip and by calculating the additional resources required to get those tasks back on track.

In terms of decision-making, AI is being used to automate the entire decision-making process while providing managers with recommendations on what actions to take, based on data and analytics, an area of AI referred to as AI-based Decision-Support Systems (AI-DSS). By way of example, AI can be used to automatically approve or reject loan applications, grant access to buildings or computer systems, or to make hiring decisions. It can also be used to identify and recommend the best course of action in complex situations where there are multiple options and outcomes, as well as with pricing decisions, product placement, target market identification and customer segmentation. AI can also assist with fraud detection, risk management, compliance, supply chain optimisation, identification of opportunities for cost savings, sourcing decisions, pricing strategies and the compilation of profit-maximising new product offerings.

In terms of control, AI is being used to monitor and control tasks, processes and systems in real time or quasi real time. It can be used to identify deviations from expected behaviour and take corrective action on predetermined or dynamically-determined rules and criteria, and has thus been prolifically put to use in myriad sectors ranging from transport and logistics all the way to agriculture, manufacturing and services. More specifically, AI can be used to optimise process performance by constantly adjusting set points and operating conditions. As such, AI can help with quality control, process optimisation, yield management, the monitoring of equipment performance, asset utilisation and employee productivity. AI is also employed in warehouses to ensure proper stock levels and to provide automated inventory management. Amazon warehouses, for instance, use robots that can identify empty shelves, optimise order queues and move inventory around.

In terms of execution, AI is being used to automate tasks that are repetitive or require a high degree of accuracy and that would normally prove to be distracting nuisances for human beings managing projects. In this ambit, AI can be used to fill out forms, extract data from documents, or generate reports. It can also be used for more complex tasks such as scheduling meeting rooms, preparing financial reports, or managing inventory levels. In terms of execution and monitoring, AI can be used to track shipments in real time, monitor production lines for quality control issues and even predict when a machine might break down.

In terms of monitoring, AI is being used to create dashboards and reports that provide real-time visibility into operations, as well as to detect anomalies and track Key Performance Indicators (KPIs) in real time. It can provide alerts when thresholds are breached or when unexpected behaviour that might require corrective action is observed. It can also help organisations understand how their processes are performing by providing insights into process flows and identifying bottlenecks and areas for improvement. It can help identify issues early on, track KPIs and compliance with Service-Level Agreement (SLAs) and monitor compliance.

It is important to keep in mind that while investments in AI are commendable, they cannot happen quickly unless the building blocks are already there. Apart from the very obvious building block of capital availability to be dedicated to AI investment, which is the sine qua non of any AI project, data availability is also important. This can usually serve purposes other than AI and is therefore an investment worth undertaking in its own right and independently of any potential future AI investment.