Dear Colleagues!  This is Asrar Qureshi’s Blog Post #662 for Pharma Veterans. Pharma Veterans welcome sharing of knowledge and wisdom by Veterans for the benefit of Community at large. Pharma Veterans Blog is published by Asrar Qureshi on  WordPress, the top blog site. Please email to for publishing your contributions here.

This blog series is based on a report from McKinsey Digital, link of which appears at the end.

Point #1 – Data is embedded in every decision, interaction, and process

The developed countries are still not there yet fully, but they often use data-driven approaches and have started using AI-driven automation for those processes which are predictive and repetitive. They still use data for other purposes where greater human input is required.

By 2025, nearly all employees are expected to routinely use data to support their work. They would also look for innovative data techniques to resolve challenges in much shorter time. Use of AI will enable organizations to leave day to day recurring activities to it, and focus on more ‘human’ domains such as innovation, collaboration, and communication. The AI shall also help improve customer service, customer experience, and use of new sophisticated applications. Data analytics and AI would be especially helpful in storage management, inventory management and procurement processes.

In Pakistan, we do make of use data for decision making, some organizations do it more often than others. However, we do it on case-to-case basis, not as a running theme even in better organization. Our second gap is that we are not even considering AI yet. It is not about money; it is about understanding and innovation. We believe more in manual labor, and do not even trust physical machines, what to talk about abstract AI. The tellers in our banks do not use counting machines because they do not trust these to do the job accurately. They count thousands of currency notes manually every day and are happy with it. Our data is also unreliable and incomplete, and we cannot use it effectively for decision making.

We shall discuss at the end the way forward for our enterprises collectively.

Point #2 – Data is processed and delivered in real time

Current technology structure and data storage practices limit data processing and analysis in real time. Even in an ERP, the data may be stored at various stations which may not be connected efficiently.

By 2025, vast networks of connected devices are expected to gather data, analyze, and deliver insights, often in real time. New technologies for this purpose shall become available which will greatly enhance the processes. Data storage shall be almost entirely shifted to the cloud computing, as its cost and availability increases greatly. Three areas/teams are likely to receive maximum benefit: Customer services/ Maintenance, Product Development, and Financial Analysts.

The benefits of real time processing are many, the biggest being the relevance with time. If we need an insight now, and get it, our decision-making accuracy and ability enhances. The teams sitting at various locations get connected easily and participate in the process. Zoom and Teams have improved connectivity, but these have nothing to do with the data processing. Other technologies are needed for fast data processing.

In Pakistan, our business landscape is very far from such ideals. We take much longer for many things without realizing it. We believe we have abundant time with us, and we are rarely hard pressed for time. For this reason, we are not looking for real time updating, processing, and analysis.

Point #3 – Flexible data stores enable integrated, ready-to-use data

Data is generated in unstructured forms and randomly. To make sense of it, it is arranged in various forms, sometimes manually, and sometimes through software. Data engineers spend considerable time exploring data sets manually, establishing relationships between them, and deriving inferences. On a smaller scale, managers also do that to understand the trends from data. However, we are referring to big data, huge amounts scattered over several platforms may be.

By 2025, data practitioners would be routinely leveraging an array of database types and will be able to organize data in more flexible ways. This would allow teams to explore and understand relationships between unstructured and semi-structured datasets easier and faster. There are several other dimensions which may be enabled through sophisticated simulations and what-if scenarios suing traditional machine-learning capabilities or more advanced techniques such as reinforcement learning.

In Pakistan, the corporate world invests huge amounts of time and money to install ERPs like Oracle, and SAP, but fails to take real benefit from it. There are multiple reasons for this failure which may be discussed elsewhere.

To be Continued……

Disclaimer. Most pictures in these blogs are taken from Google Images which does not show anyone’s copyright claim. However, if any such claim is presented, we shall remove the image with suitable regrets.

Leave a Reply

%d bloggers like this: