No matter big or small, companies around the world have been experimenting with AI to stay competitive in their respective markets and explore the benefits delivered by digital transformation. According to Harvard Business Review, Data Scientist is the Sexiest Job of the century!
We caught up with Marta Markiewicz and Michał Zgrzywa, Heads of Data Science at Objectivity:



Thank you for taking your time to meet me. I can tell that you guys are very passionate about Data Science– the way you talk about it and describe your team and projects you’ve worked on. 

We all have heard about the impact AI is having in both business and daily life, what trends are you seeing in AI & Data Science, any favourites?

MM: There are so many trends I find fascinating and captivating that it’s hard to name just one. I’m seeing exciting things happening in feature engineering, reinforcement learning and data privacy. I also anticipate potential in digital twins – virtual mirrors of reality which enable you to create, build and test processes and services in a parallel, virtual world. 

MZ: For me, autonomous vehicles are the most fascinating outcome of the AI revolution. When Arthur C. Clarke said “Any sufficiently advanced technology is indistinguishable from magic” I think he meant applications such as these. It is truly remarkable to see multiple AI and engineering solutions combined in the final form of a car or a robot. And we are still at the beginning of this journey.

There is no doubt that like never before, we have access to so much data (by 2025 we should have 180 zettabytes of digital data); and there is a huge potential in the data to be unlocked. Which is the bigger challenge for an organisation: not having enough data to work with or having too much unorganised data? 

MM: For me, there is a third option: having dirty data, whatever the volume! Whenever lacking enough data, an organisation can always start collecting it, from that moment onward. Also, when thinking broadly enough, there is often useful data already collected, just not the obvious source. In my opinion, there is no such thing as too much data and increasingly the value in unorganised data can also be harnessed, just like the organised data. However, what is usually a true challenge, is the dirt in the data. Analysing the data over 8 years, I’ve literally NEVER seen a perfectly clean dataset. And the worst examples provide a false sense of security data sets that seem clean but actually aren’tthey can lead you into a dark alley of misleading insights. 

Sounds like data very easily can give us a headache. Not long ago we hosted a Forrester Analyst Day here at Objectivity. We learned that AI gives enterprises the power to predict and that data-driven businesses perform better. What challenges are holding back companies from achieving their potential using data science?

MM: Data quality (see the answer above) and distrust mixed with anxiety. Automation of tedious tasks, which can be achieved with data science, can be perceived as a job stealer, instead of a facilitator. As a consequence, employees can be reluctant to share their know-how. What’s more, data science projects are risky, they may be quite expensive and at the same time the precise process and absolute outcomes can be difficult to articulate. Using complicated models, not entirely understood by users, can make it difficult to gain trust. This can lead to data science projects being classified as non-crucial parts of the business, not viewed as true game changers, just providing interesting insights. 

MZ: The biggest blocker in AI adoption across the companies is the struggle to innovate. AI projects tend to be technologically challenging, risky, and difficult do deploy. When such projects set out to decrease a small portion of cost in a non-crucial business process, the ROI is not great. This discourages people from trying. But when a project stems from an innovative idea to reinvent the business operations or create a new line of business, the result can be astonishing and worth the risk. It can be difficult to innovate with a day-to-day business challenges and lack of awareness of what the technology is really capable of. 

Some of our biggest clients represent retail and we do observe that AI is accelerating in this industry especially fast; in what other industries do you see opportunities for AI and Data Science? 

MM: Actually, I see it literally everywhere! My favourite domains include manufacturing and agriculture. One of the initiatives we carried out last year was from the manufacturing industry; people had to check items on production line manually, which is an extremely tedious and fault-prone procedure. However, with AI incorporated into the production line, the items could be checked with only minimal human intervention.  As for agriculture, I’m fascinated with the following possibilities: drones flying over fields to identify areas requiring attention and automatically watering crops. 

MZ: For me, the biggest opportunity lies not necessarily in a specific industry but relates to specific tasks people perform at work. I think of an analogy to the industrial revolution, where the invention of the steam engine enabled machines to perform simple physical tasks at a low cost. If someone was running a factory back then and was able to imagine another way of performing simple tasks like pushing, moving or lifting, the steam engine allowed them to innovate and disrupt their markets. I believe the AI revolution will be similar. We have a possibility to perform simple tasks, that so far were reserved for humans (image recognition, understanding the message in a text, spotting a trend or correlation in a series of numbers), by applying much cheaper solutions. Companies that are able to re-imagine their businesses to use this opportunity, will disrupt their industry and win market share.        

Good that you mentioned the tedious tasks that machines could help us out with.  Do you think Data Science tasks currently done by humans will be automated in the near future? That’s definitely one of the most common fears, that robots will steal our jobs. 

MM: Automating tasks, yes, especially the tedious ones; which is actually very good thing – people could finally focus on their creativity! 

MZ: There are many kinds of Data Science tasks. There are purely technical tasks like data cleaning, choosing model parameters, preparing simple visualisations. There are tasks that require broad knowledge and creativity like choosing a proper model or feature engineering. Finally, there are tasks focused on understanding the data and business around them, communication with stakeholders, storytelling and other that require soft skills. I expect automation mostly in the first group. It should be very welcome – people could use their intellect to solve more challenging tasks. 

You mentioned that Artificial intelligence is everywhere; I am curious how you guys work with it on a daily basis. Could you tell us a bit about your team at Objectivity and maybe some projects you work on?  

MM: With pleasure! My team consists of 12 people, experts in data crunching, modelling and visualisation. The team has a varied background in mathematics, statistics, computer science, physics, engineering, economics, neuroscience, law. Every day we solve challenging business problems combining the use of domain knowledge with sophisticated algorithms, statistics, and machine learning.  

As for the projects, the most interesting one right now is about automatic suggestions for email responses. IT Support receives a lot of requests through emails and a lot of them are recurring. Semi-automatic responses would speed up the process and unburden IT guys from repetitive, tedious tasks. If you would like to learn more, there is my article describing Objectivity approach and cooperation with our clients in the recent issue of The Sunday Times.

Every technology revolution causes some social uncertainty; Michal, you mentioned the industrial revolution and how people were struggling with it at first. Will it be the same with AI and Data Science?

MZ: We all have read in the news about the cases of poor AI application (the problem with a recruitment algorithm that showed gender bias or self-driving cars having accidents). People are also afraid of losing their privacy; they are not comfortable with companies gathering too much information about them. Finally, there is much uncertainty around jobs – will machines push us towards unemployment so there is anxiety, some of which is justified.  

We definitely need to introduce new regulations, GDPR was the first step and there will be more e.g. regarding bias of AI models or regarding legal responsibility of AI actions. But this revolution will not be stopped and the change will come. If applied responsibly, society will be able to adapt – it is worth bearing in mind that the industrial revolution removed jobs in the short term but created a lot of new ones in the long run. It also pushed humankind into developing even further opportunities. It is likely that AI will do the same. 

We hope you enjoyed this interview and don’t hesitate to get in touch if you have any questions related to AI & Data Science 


Marta Markiewicz
Head of Data Science at Objectivity 

Marta is a data scientist with over 7 years of research experience in Artificial Intelligence and Data Science. She graduated of Mathematics at the University of Technology in Wroclaw. She has worked as an Analyst as well as a Developer. She has carried out foundational work in computational models using machine learning techniques. Together with her team she has helped built intelligent systems. In 2014, she became a Data Scientist at Objectivity. In 2016 she became a Team Leader and shortly after, Head of Data Science department. She is a conference speaker and occasional writer (Sunday Times, Programista). 

Michał Zgrzywa
Head of Operations Research at Objectivity 

Experienced leader of software delivery departments and organisations. In the past helped building small-to-medium-sized businesses taking on many roles, depending on the needs of growing organisations. Big supporter of motivation through autonomy, trust and responsibility. Always up to date with technology. Currently involved in the development of Data Science and Artificial Intelligence practices at Objectivity.