Meet Katherine Munro – Artificial Intelligence: The Future of Marketing

Meet Katherine Munro - Artificial Intelligence: The Future of Marketing

There is no doubt that AI is revolutionizing the world of marketing. This is helping retailers to provide more assistive, enjoyable experiences for customers, and improving performance with it.

It’s called ‘predictive personalization, and it’s one of the hottest trends in AI-driven marketing.

One-to-one messaging that doesn’t sound like spam, dynamic website design that adapts to individual users, and retail stores that recognize and welcome their visitors.

Katherine is a Data Scientist at Smarter Ecommerce GmbH. She is responsible for building AI and machine-learning driven solutions for e-commerce. As a Data Science Ambassador there, she conducts training in AI, ML, and data science. With a background in computational linguistics and (deep) machine learning, Katherine has worked in research and development for Mercedes-Benz and the Fraunhofer Institute, specializing in UI and natural language understanding.

That’s why on October 19th 2020, we had a free Data Science webinar, “AI in Marketing” where Katherine shared her own insights and experiences about the most advanced technology to date.

Brainster: Please, introduce yourself to the Brainster Audience. Tell us more about your career. What made you pursue a career in Data Science?

Katherine: A fairly winding career path I’ve had to get here. I worked in finance while studying business and graphic design, and eventually stumbled into linguistics – the scientific study of all of the world’s languages. Even I fell in love with it,  by the end of my Bachelors’s Degree, I was searching for a way to combine it with the business and creative aspects of my past.

I decided to move from Australia to Germany to specialize in computational linguistics, which introduced programming and Artificial Intelligence into the mix. I loved the focus on creatively solving real-world problems, like how to improve translation services, smart home assistants, and other technologies we use every day.

As a student, I did R&D for Mercedes-Benz and the Fraunhofer Research Institute. I got to use computers to dig into language data, and then machine learning and neural networks to derive new solutions. It was brilliant. When I saw the ad for a Data Science position, I realized I could keep doing all of this, while being creative and driving real business value. I knew it was the job for me.

Meet Katherine Munro - Artificial Intelligence: The Future of Marketing

Brainster: On October 19 2020, we had a free webinar titled “AI in Marketing”. What can viewers learn from it?

Katherine: As I’m sure you know, AI is revolutionizing many of our daily habits and interactions, including how we shop. Today’s consumers are time-poor, have many demands on their attention. They have been trained by the likes of Google and Amazon to get exactly what they want, right now. They are demanding that any marketing interactions they have with brands be personalized and relevant. Not only does that improve their experience; it can help businesses build their brand reputation and increase revenues.

But this is only possible via AI techniques that can process massive amounts of behavioural data and derive insights on exactly what consumers want. My talk shows what kinds of AI-driven personalized marketing are now possible, and go through some real-world examples.

Brainster: As a professional, you might be tired of the hype surrounding Data Science, but as we can see that hype is going nowhere. Did you expect this to happen though? How Data Science became the “hottest job” of the 21-st century?

Katherine: Firstly, I’m not quite tired of it yet – it’s still nice to hear ‘wow, cool!’ when I tell someone what I do. And as my first answer shows, I never expected to be here.

But the reasons for the demand are clear. In the early days of AI, research was limited by available data and compute power. For example, as early as the 1950’s the first neural network was applied to a real, but necessarily simple, problem.

But as computers got better and more data began to be collected, at a pace that is only increasing thanks to social media and the Internet of Things, AI started to achieve more. And the more it can theoretically do, the more companies are willing to invest in trying to convert pure academic work into business solutions.

This is where the hype you mentioned can become problematic, as not all problems require a complicated, AI-based solution. But as long as we have a healthy mix of creative, big-dreamers, and data scientists who know the genuine strengths and appropriateness of various AI techniques, I believe the field will continue to thrive.

Brainster: What do you think is the one and the most important prerequisite for a person to start studying Data Science? Strong knowledge of maths and stats? Any sort of coding skills? Excel? Experience in working with datasets? (This answer is pretty much combined with the next question)

Katherine: There’s no doubt one should be comfortable with concepts like descriptive statistics and probability. Luckily, there are wonderful resources available to help even those who haven’t thought about maths since high school gets themselves skilled up in preparation.

Hopefully, the data science course will teach you programming, but if it doesn’t, there are again resources available. And any work with Excel and datasets will help, purely as a primer for proper data science training. But some less obvious traits are even more important: the curiosity to go digging into data, the ability to be delighted by unexpected patterns within it, the creativity to come up with explanations as to what’s going on, and the determination to then go hunting for the real answer.

Meet Katherine Munro - Artificial Intelligence: The Future of Marketing

Brainster: In your experience as a lecturer, which part of Data Science is the hardest to master?

Katherine: The hardest part really depends on personal characteristics. For example, whether someone loves maths or has already been a hobby programmer all their lives. Clearly, this will affect which subjects they find easiest. Just as someone who loves theory but hates repetition might find it easier to read a machine learning paper than try to implement the algorithm for themselves.

I think it’s best to approach a data science study being grateful that it covers such a diverse range of activities. So there will always be something you love and can specialize in. As for the fear of coding, a good way to tackle this is to teach it in a way that starts small, showing that code can be logical and easy to read, and build up from there, to show that even the most complex programs consist of smaller, easier-to-define chunks.

Secondly, we should let students play with code and try to build whatever they want. They’ll probably surprise themselves (and the teacher!) with what they can come up with. When it comes time to show them something more complex, if you can frame it as a tip for how they can improve what they’ve already built, they’ll welcome it and be able to immediately contextualize it.

Brainster: COVID-19 increased the pace with which the digital transformation is taking over the world. AI is slowly integrating into all business sectors and is becoming a “staple” technology. With this, the media is slowly pushing the narrative that soon AI will be able to solve and prevent major adversities like natural disasters or pandemics. Do you agree with this statement? If so, how soon is “soon”?

Katherine: You saved the hardest question until last! The thing is, I’m an obsessive follower of the latest research and the greatest minds in AI. They all admit that they don’t know if or when the next big breakthroughs might arrive, and what they will look like. That’s because we’ve been wrong in the past: the two AI winters arose because AI wasn’t able to outperform the hype around it, whereas the rapid advances in neural learning brought about by GPU computing took everyone by surprise.

Currently, even the best AI is still ‘narrow’. Meaning it is brilliant at one or a couple of purposes, but nothing broader. That’s why humans will still be driving solutions for now. As they’ll be able to take the insights delivered by the best AI and contextualize them in the real world where the real problems are. Some experts believe the biggest challenges, like climate change and getting humans to Mars, are so complex that they can only be solved by AI. Who knows? Let’s just keep a timestamp on this interview, and you can ask me again in five years!

Watch the recording of our webinar “AI in marketing”, with Katherine, as a guest lecturer.

If you have ever thought about a future-proof career in Data Science, have a look at our remote Bootcamps

What do you want to be when you grow up?