Connect with us

Interviews

Do Interviews 3: Paul Kennedy

Data Overflow

Published

on

[simplicity-save-for-later]
Our third interview was with Paul Kennedy, science communicator for more than 5 years in Cape Town, South Africa, project manager, and now occupying the position of community and communication manager at Zindi Africa. In this interview, we’ll be talking about UmojaHack Tunisia 2020 and UmojaHack Africa 2020.

Q:
Introduce yourself with your own words.

A:
I was very passionate and enthusiastic about science and innovation and I thought at university that I would want to be a researcher that I would be very happy in a lab working on difficult problems and try to make the world a better place, and it took for me to actually complete research of master to realize maybe I’m not actually so well cut out to be a research scientist it takes a very specific sort of person to be a researcher they pay attention to details, and sometimes it’s quite learnly position to be working in a lab so at that point I realized i was still very passionate and enthusiastic about science and about it wasn’t going to work for me to be a researcher so then I also at the same time I started thinking how I could still be involved in this landscape I guess, in the science landscape and I started to look into science writing and science communication, I’ve always enjoyed writing and I thought that might be a good way to combine these two things then I moved into writing about science, so very quickly met somebody who was doing something quite similar to me, she has a research background also realized she didn’t really want to be in that world, but she enjoyed talking about science and innovation, and so we together formed a business working on science and communication we kinda did that for about five years, on that time we didn’t write just about science, but we also worked in a lot of I guess you can call it multimedia, we worked at data visualization, we worked on building websites, and we built several applications then from there when I saw the position opened up at Zindi I realized that it was quite an unusual role, but it was well suited to my skillet of understanding the technical side of things but also being able to talk to people and knowing how to communicate quite complicated concepts easily, so I joined zindi a year ago.

Q:
Mr. Paul, you studied oncology and cancer biology then you got a master’s degree in biotechnology. This is a very sensitive field that keeps evolving and requires a lot of knowledge and precision. So how did digitalization affect your career? Did it make it any easier with all the new technologies?

A:
I think the biggest way it affected my career is that it gave me new opportunities, I think during my studies it hadn’t much affect but when I started to work in science writing and communication I realized that all these amazing tools would help, I learned how to use online tools, and it represented an opportunity for me, I taught myself how to blog, I started to blog myself and that taught me a little about how to use
content management system, I taught myself how to use social media to communicate. The whole process of digitalization was really about opportunities and ways to learn new things outside the academic environment.

Q:
Now, we would like to talk a bit about Zindi Africa. Can you introduce it to us?

A:
For me, I think the best way to think about Zindi is that it is this phenomenal Community for data scientists in Africa. There are people all over Africa looking to develop careers in data science, and Zindi is the place where they can come to do that, to meet other people like them to develop new skills to learn new things to sharpen the practical guide to things and hopefully to get a job as well in Data Science in
Africa. We recently went past 20,000 users something like three-quarters of that are in Africa and Tunisia is our third or fourth biggest region, and it is really just a platform that is aiming to revolutionize data science in Africa and not for anybody else is Africans doing it for Africa.

Q:
As you said, Zindi is the first platform for data science competitions in Africa, as a communication expert what is the importance of participating in hackathons?

A:
I think you can learn almost anything in theory online there are many courses available a lot of different materials videos, blogs, tutorials, and it’s all out there. But what there is a lack of somebody trying to break in today’s data science as a career is practical experience so to me a competition platform like Zindi is a place where people can take something that they learned and put it into practice to see how it works in reality because it is one thing to understand a practical skill like data science its one thing to understand the theory of it, and it’s entirely different to try and put it into practice in a real world problem. So for me, it’s a place where people can learn to apply their skills more practically, it’s a place where people can gain confidence in applying their skills and I think that’s the real value of competition.

Q:
Is this platform available for every hackathon organizer in Africa, or are there certain required criteria and conditions to be able to host a competition on Zindi’s platform? And what steps should be taken?

A:
As I said we want to be THE community for data science and for that reason any community involved in data science who would want to run a hackathon with Zindi can come to us and do so. We encourage that, and we invite everybody to give us a call and see what we can do together we feel very strongly about collaborations and about community building it’s also a place other organizations say businesses and startups also can come to us to run hackathons and competitions as I said we are very open to almost any organization approaching us we can almost certainly find one way or another to collaborate. So if you want to host a competition with Zindi it depends on if you have data or not? So Zindi hosts long-term competitions as well as hackathons and if you have a problem, and you have a dataset that you would like to try and use it to solve that problem you can bring that to Zindi, together we can work on it. Or even when you do have a dataset, and you’re not sure what to you could do with that the come to Zindi. With organization, we will try to understand the problem better to see how we can shape it into a competition and to post that competition alternatively if you are on the community side you don’t necessarily have a dataset, but you have a group of students of learners of data science enthusiasts, and you want to use a hackathon to say turn up excitement about data science in your community or perhaps add a practical element to a course a workshop a boot camp that you are running then you can approach us, you can email me or reach us on social media then we will usually draw data from one of our other competitions we will replicate that competition for a community to run their hackathon.

Q:
It’s a known fact that the biggest problem that can occur in competitions based on Data Science and AI is the lack of data. Did you face this issue while organizing your events? If yes, how did you deal with it?

A:
I think I can safely say we haven’t faced that problem. Zindi has been around for more than 2 years now from our perspective we’re proving that it’s a myth it’s not actually true that there isn’t data in Africa. We’ve run almost a 100 competitions more than 60 datasets in the course of the 2 years and I think that’s really one of the powerful things about Zindi that we are enabling those datasets to come out of hiding and to be used and to be put into practice for the good of everybody for both people who want to learn data science but also for people living in Africa affected by any number of problems I think that’s one of Zindi‘s great strength

Q:
UmojaHack Tunisia 2020, The objective of the challenge was to predict the class of a protein kinase enzyme using only its amino acid sequence. According to what needs to be detected in Africa was the theme chosen?

A:
I think previously one of the things that we really tried hard to do is to be in a « bling track form » and for us, this competition was about innovation the challenge is something that never been done anywhere in the world never mind anywhere in Africa, and we want to show that Africa can take the lead on it’s fighting innovative deep learning projects like this and the hackathon challenge was for our partner Instadeep, Instadeep are world leaders in AI and deep learning, and they are coming to Africa for solutions. It’s maybe not the most stressing need, but I think it’s showing what Africa can do and showing that we can compete with the rest of the world in that field.

Q:
It’s a known fact that the biggest problem that can occur in competitions based on Data Science and AI is the lack of data. Did you face this issue while organizing your events? if yes, how did you deal with it?

A:
Well, I think that we haven’t faced that problem yet, technically for around two years now, and from our perspective, we are proving that it’s not actually true that there isn’t data in Africa. We have more than 100 competition, more than 60 days to set for the course of two years and I think that the powerful thing about Zindi is that we are enabling amateurs to come out from hiding and to the use of practice for the good of everybody, both for people wanting to learn and for people in Africa affected by any number problems.

Q:
UmojaHack Africa 2020, the first virtual inter-university machine learning hackathon in Africa. We are talking about 2500 students from 70 universities in 17 different countries. Can you tell us about the course of this competition?

A:
We envisioned UmojaHack Africa as something to unite people around countries, in fact, Umoja means unity that’s why we chose the name, and we tried to find challenges that would be relevant and accessible to students at every levels and we were thrilled about them we got much more participants than we expected.

Q:
Based on your own experience, what are the things you believe a hackathon organizer should do to guarantee the success of his event?

A:
I think a good dataset is probably the place to start because for a hackathon to be successful, the participants really need to be invested and enthusiastic and excited about working on the competition, so a good relevant dataset is so important and when we run hackathons in a specific country, we try our best to get a dataset from that country. As in the case for UmojaHack Tunisia, that is a dataset from InstaDeep which is a Tunisian company. So we try to make the dataset good and relevant. And the platform of course is very important. I think a lot of data science hackathons couldn’t happen in Africa without a platform like Zindi where we can objectively score submissions in quite a short time, that is important. And I would say checking the data really carefully and checking it more than once. We put our dataset through a really regular review process because there is a lot that can go wrong in running a competition in terms of data leaks and maybe not having the data format in the correct way. And then lastly I think a good marketing plan or strategy to make sure people know about the event, make sure there’s some support from sponsors, and make sure that there is enough enthusiasm and enough time for people to sign up for the event and looking at ways to reduce various entries.

Q:
Now, we would like to know what you think of the current state of AI and Data Science in Tunisia, and Africa in general? We all know that we’re new to the field compared to other countries that set the fundamentals of this science, but we can tell that the interest in Data Science is increasing significantly lately in Tunisia. So, we want to ask you what are the obstacles holding us back according to you?

A:
I must give a clarification that I am new in this field. I’ve been working at Zindi for a year. Before that time, I couldn’t tell you what a regression model was, I didn’t understand even the smallest things about data science, so I’m speaking as an outsider. My point of view is really just incredible revolution in Africa around Data science and what’s really amazing to me that it’s driven from the ground of its people. I mean people like you who are just so passionate and enthusiastic about it, and you’re not going to let circumstances, money, and other limitations hold you back from learning data science and using it to make your environment and also your great environment all the way up to Africa and the world a better place. And I think that’s probably, to me at least, what sets Africa aside in terms of AI and data science is this incredible passion from the communities all over the place. So I think that’s really amazing. I can’t speak about Tunisia particularly, I know that Tunisia is one of our strongest countries. And I think there is quite a good support for data science from universities. I think that makes a big difference. Tunisia was I think our largest UmojaHack. I have to check but I think it was our largest UmojaHack. We had more than 250 people, and we were really happy that we had 28% of participants as women and on our platform, we have 23% women which I am already very proud of, but to see such enthusiastic women on the competition from Tunisia was really amazing. So I think there have been so many barriers in Africa, but it really seems like things are changing so incredibly rapidly in this field of data science, and it’s all about the people.

Q:
What are your hopes and aspirations for Africa? Where do you see it in 20 years?

A:
It is very hard to make such statements. But seeing the uptaking enthusiasm for data science and seeing how Africa data scientists are now being employed by organizations all over the world with making an impact all over the world, I think it is only going to improve. I think there is a really remarkable progress. It looks to me like a revolution. We will just keep improving that for the next years.

Q:
As a community and communication manager, what strategy do you follow to keep your team motivated?

A:
I work with really incredible investors. We have 54 investors across Africa from 26 countries. I don’t feel like I’m leading that team. I feel like I’m just a part of an amazing group of people. They are all volunteers, they all work in their own communities, they are incredibly passionate, skilled, hardworking people. So I don’t feel like I need to do very much. I try to encourage people to talk to each other. I try to be a good instructor. If they need something I try to be able to give that to them. I would love to finance every community in Africa. That would be amazing. But we can’t do that. So we try to find other financial ways to support them. I feel that I’m a part of an amazing team. I don’t feel I’m leading them. I don’t need to tell them what to do. They come to me with ideas more often than I come to them. So I’m very lucky!

 

Share your thoughts

Continue Reading

Interviews

Beyond INSAT: Tech Giants | Episode 1 : Dr. Skander Hannachi

insatpress

Published

on

[simplicity-save-for-later]

By

  • Guest: Dr. Skander Hannachi
  • INSAT Class: IMI 2003
  • Publishing Date: Monday May, 17th, 2021

Our first guest in the “Beyond INSAT” podcast is Skander Hannachi. Currently an AI / Machine Learning specialist at Google cloud; Dr. Hannachi, in full transparency, speaks of INSAT, family, hobbies, the different positions he held in his respective field (Artificial Intelligence / Machine Learning), and more. He talks about his INSAT journey, being part of the very first class of INSAT’s students, his passion for music, and gives invaluable advice to INSAT students and college students in general. He clears many misconceptions about post-college life through his own experiences and objectively addresses the potential Tunisia has in the Tech Industry.

Listen to Mr. Hannachi’s interview with Eya Zaoui now! Available on insatpress.tn and Soundcloud.

  • Book Recommendations by Mr. Hannachi:
    • The Structure of Scientific Revolutions (1962) by Thomas Kuhn.
    • The Open Society and Its Enemies (1945) by Karl Popper.

Realised by:
Eya Zaoui & Majd Akkari.
This work wouldn’t have been possible without the efforts of the INSAT Press team.
Special Thanks go to Skander Soltane, INSAT Press’s Editior-in-Chief and this work’s fabulous supervisor!

Share your thoughts

Continue Reading

Made with ❤ at INSAT - Copyrights © 2019, Insat Press