Oguzhan Gencoglu, topdatascience.com
Hi Oguzhan, could you briefly introduce yourself to our readers?
Absolutely! My name is Oguzhan Gencoglu and I am the Co-founder and Head of AI at Top Data Science. My role is to lead machine learning development and provide guidance to our data scientists. Besides, I also implement machine learning and deep learning solutions on a daily basis. I have a strong background in ML research. So far, I have conducted ML research in several countries, including Turkey, USA, Denmark, Czech Republic, and Finland.
Impressive! So, you are both a traveler and a cross-continent researcher. Pretty cool! Now, as you know, data science became quite popular relatively recently. When was the first time you heard about the field and how did you end up working in it?
Well, I was introduced to data science during my BSc studies 11 years ago, particularly from the signal processing and statistical estimation point of view. In addition, I took relevant courses throughout my BSc, MSc, and PhD Studies. Those included pattern recognition, machine learning, estimation theory, probability theory, etc.
It genuinely seems that you really made the most of your studies, doesn’t it – that’s quite inspiring. As one of the founders of Top Data Science, could you tell us a bit more about the services you are offering?
Sure! We are Top Data Science and we provide AI-as-a-Service. Our company has delivered 60+ machine learning solutions in 2 years to our clients throughout the world. Furthermore, Top Data Science offers machine learning and deep learning development services along several industries. To provide these, we employ our joint expertise. We are a high-quality team of 12 AI experts, and half of us have PhDs.
Sounds great, Oguzhan! Could you tell our readers when you started your business venture?
I met my 2 other co-founders in 2015. We officially started Top Data Science in April 2016.
They say great minds think alike. We’re positive that has a lot to do with your success. Nevertheless, could you shed some more light on how you create value for businesses?
We create value to our clients in several ways, such as automating processes, speeding up decision making, saving costs etc.
Awesome! And which area of improvement is typically the easiest to address? More specifically, which area allows you to achieve some quick wins when you start working with a new company?
To tell you the truth, the bulk improvement comes when the client understands the main concepts of machine learning and, therefore, has already prepared relevant data for training algorithms. Thus, we get extremely quick results, especially in the computer vision domain.
Interesting! Why do you believe this is the case?
To a large extent, I believe this is due to our deep expertise and long history in image and video data.
There is no doubt Top Data Science has a lot to offer to its clients. Out of curiosity, which industries do you typically work in? Would you say all data is equal or is there a particular industry you consider more challenging?
Well, we work in several industries including healthcare, forestry, construction, life sciences, and the automotive industry. In general, healthcare data is more difficult to operate with, due to privacy concerns. That’s especially true if the data is coming from individuals or patients.
That makes sense, for sure. In terms of data processing, which tools and software are essential for your team?
We develop mostly in Python and scientific libraries such as numpy, scipy, and pandas are essential. When it comes to conventional ML, we use scikit-learn/xgboost/GPy. Continuing the list, nltk and gensim are essential for our NLP projects. Naturally, for deep learning, we prototype in keras/TensorFlow and deploy in TensorFlow, if it is an industrial collaboration. We have also used PyTorch for research collaborations. Of course, TensorBoard and jupyter notebooks are also widely used… SQL and relevant libraries for database interaction are also part of our daily-used tools.
I bet our readers will be super happy with this list – thanks for the detailed info! Now, can you think of a situation when you have worked with a given company and you felt particularly proud of what you helped them achieve?
Yes, absolutely! What comes to my mind is our collaboration with the Helsinki University Hospital. We developed AI algorithms for detecting prostate cancer and we have achieved outstanding results. We even published scientific papers together. I am quite proud of this work.
You can find more details here: https://arxiv.org/abs/1903.05769.
And a video here: https://twitter.com/TopDataScience/status/1049316899090309121
That’s so inspiring, Oguzhan! We’re completely blown away! You’ve already given our readers tons of helpful insights, but we still need to ask: tip of the interview? Is there a nifty tool that you discovered or were introduced to, which you now can’t live without and want to share with an audience of aspiring data scientists?
Sure! I nowadays enjoy dash by Plotly. It’s a pretty nifty tool for quick interactive demos. It’s kind of similar to RShiny in R. Plus, it is much better for showing interactive visualizations to the client, compared to powerpoint.
Every time we conduct these interviews, we finish with some nerdery. What is the one nerdy thing you would like to share with the world? It doesn’t have to be data science related.
One nerdy thing I enjoy is called Secretary Problem in mathematics. Basically, if you are planning to hire for a given position and there are N applicants, the optimal stopping rule prescribes always rejecting the first N/e applicants that are interviewed and then stopping at the first applicant who is better than every applicant interviewed so far.