DL Indaba: AI Investments in Africa

This week we are bringing you a couple of interviews from last week’s Deep Learning Indaba conference. Dr. Vukosi Marivate, Andrea Bohmert and Yasin(i) Musa Ayami talk about the burgeoning machine learning community, research, companies and AI investment landscape in Africa. While Mark is at Google Cloud Next in Tokyo, Melanie is joined by special guest co-hosts Nyalleng Moorosi and Willie Brink.

Vukosi and Yasin(i) share how Deep Learning Indaba is playing an important role to recognize and grow machine learning research and companies on the African continent. We also discuss Yasin(i)’s prototyped app, Tukuka, and how it won the Maathai Award which is given to individuals who are a positive force for change. Tukuka is being built to aid economically disadvantaged women in Zambia get access to financial resources that are currently unavailable. Andrea rounds up the interviews by giving us a VC perspective on the AI start-up landscape in Africa and how that compares to other parts of the world. As Nyalleng says at the end, AI is happening in Africa and has great potential for impact.

Willie Brink

Willie Brink is a senior lecturer of Applied Mathematics in the Department of Mathematical Sciences at Stellenbosch University, South Africa. He teaches various courses in Applied Mathematics and Computer Science, at all levels, and his research interests fall mainly in the broad fields of computer vision and machine learning. He has worked on multi-view geometry, visual odometry, recognition and tracking, probabilistic graphical models, as well as deep learning. Recent research directions include visual knowledge representation and reasoning. Willie is also one of the founders and organisers of the Deep Learning Indaba, an exciting initiative working to celebrate and strengthen machine learning and artificial intelligence research in Africa, and to promote diversity and transformation in these fields.

Nyalleng Moorosi

Nyalleng is a Software Engineer and Researcher with the Google AI team in Ghana. Before joining Google, Nyalleng was a senior Data Science researcher at South Africa’s national science lab, Council for Scientific and Industrial Research (CSIR), with the Modeling and Digital Sciences Unit. In her capacity at CSIR, she works on projects ranging from: rhino poaching prevention with park rangers, working with news outlets to understand social media sentiments, and searching for Biomarkers in African cancer proteomes. Before getting into ML research at CSIR, she was a computer science lecturer at Fort Hare University and a software engineer at Thomson Reuters. Moorosi is an active member of Women in Machine Learning, Black in Artificial Intelligence, and an organising member of the Deep Learning Indaba - a yearly workshop that gathers African researchers in one space to share ideas and grow machine learning and artificial intelligence capabilities.

Dr. Vukosi Marivate

Dr. Vukosi Marivate holds a PhD in Computer Science (Rutgers University) and MSc & BSc in Electrical Engineering (Wits University). He has recently started at the University of Pretoria as the ABSA Chair of Data Science. Vukosi works on developing Machine Learning/Artificial Intelligence methods to extract insights from data. A large part of his work over the last few years has been in the intersection of Machine Learning and Natural Language Processing (due to the abundance of text data and need to extract insights). As part of his vision for the ABSA Data Science chair, Vukosi is interested in Data Science for Social Impact, using local challenges as a springboard for research. In this area Vukosi has worked on projects in science, energy, public safety and utilities. Vukosi is an organizer of the Deep Learning Indaba, the largest Machine Learning/Artificial Intelligence workshop on the African continent, aiming to strengthen African Machine Learning. He is passionate about developing young talent, supervising MSc and PhD students, and mentoring budding Data Scientists.

Yasin(i) Musa Ayami

Yasin(i) Musa Ayami is Team Lead at TsogoloTech and a certified Oracle Associate. Mr. Ayami recently graduated with a Master’s Degree in Information Technology at the prestigious Durban University of Technology (DUT) were his study mainly focused on Computer Vision and Machine Learning. Prior to him enrolling for his Master’s Degree, Mr Ayami served as an Intern Software Engineer at DUT’s App Factory where he also served as Team Lead before deciding to further his studies. He also worked as a Part-Time Student Instructor at the DUT. In 2017, he co-founded TsogoloTech. His vision has always been to leverage technology for social good.

Andrea Bohmert

Andrea Bohmert is a Co-Managing Partner at Knife Capital. Before joining Knife Capital, she was the Founder and Co-Managing Partner of Hasso Plattner Ventures Africa. Passionate about strategizing how to scale businesses and meeting the entrepreneurs responsible for creating them, she has been actively involved in numerous initiatives aiming to accelerate the African entrepreneurial ecosystem.

What are you looking forward to this week?
  • AlphaGo Movie site
  • WiML: Women in Machine Learning site
  • Deep Learning Indaba Poster Sessions site
  • Neural Information Processing Systems site
  • Deep Learning Indaba site
  • Deep Learning Indaba GitHub site
  • Deep Learning Indaba Tutorials site
  • Deep Learning Indaba 2018 Slides site
  • Deep Learning Indaba 2017 Presentations videos
  • Deep Learning Indaba X site
  • Yasin(i) Musa Ayami on GitHub site and LinkedIn site
  • Deep Learning Indaba Award Winners site and tweet
  • Maathai Award site
  • Xamarin site
  • SuperPosition at The Deep Learning Indaba with Dr. Vukosi Marivate podcast
  • Knife Capital site
  • Investing in AI by Andrea Bohmert article
  • 10 Defining Moments that shaped the 2016 SA startup ecosystem article
  • Data Science Africa site
  • International Data Week site
  • Google Cloud Platform Credits award winners tweet
Question of the week

The co-hosts weigh in on our question of the week: What have you taken away from this week and will take forward?

Where can you find us next?

Mark and Melanie will be at Strangeloop.

Willie will be teaching Machine Learning at Stellenbosch University this summer.

Nyalleng will be at the Women in Machine Learning Workshop and the Neural Information Processing Systems Conference in Montreal in December.

[MUSIC PLAYING] MELANIE: Hi, and welcome to episode number 147 of the weekly Google Cloud Platform podcast. I'm Melanie Warrick, and I'm here with special co-hosts Willie Brink and Nyalleng Moorosi. Hello, Willie. Hello, Nyalleng.

NYALLENG: Hi, Melanie.


NYALLENG: It's good to be here.

MELANIE: Thank you, for being here. All right, I'm going to have you both take a quick minute and tell us a little bit about who you are.

WILLIE: Thanks. I'm Willie. I am a lecturer at Stellenbosch University. This is, of course, also the host for this year's Deep Learning Indaba. I have a passion for machine learning, computer vision, I have a passion for teaching, and it is wonderful for me to be here.

MELANIE: Well, I'm glad you're here. Nyalleng?

NYALLENG: Yeah, I'm Nyalleng Moorosi, as Melanie has mentioned. I recently joined the Google AI team for the group that's going to be Play. That's going to be in Accra, Ghana. I'm very excited about that. Before that, I was a researcher at the Council for Scientific and Industrial Research. It's a very, very long name, but it's a research lab here in South Africa, government funded, and also partially industry. I'm very, very excited to be going to Accra, because I can continue to do machine learning and continue living in Africa, which is exactly where we're here at the Indaba.

MELANIE: Yes, we are doing this recording, and this episode specifically is about Deep Learning Indaba where we're going to showcase a couple of interviews that we captured during this week. Can you both take a minute and just tell the audience what Deep Learning Indaba is about?

NYALLENG: Our line is strengthening African machine learning, so it's always good to start with that. But I think the Deep Learning Indaba has become, strengthening African machine learning. But I think when we started it, we did not know exactly how much it would. And it has become a community of sharing knowledge, of sharing skill, of celebrating one another and, of course, continue on learning. So it almost feels like it's going to be our yearly pilgrimage where, from wherever we come from, we go to wherever everybody is and we collaborate, and we work together, and we just-- we catch up and we celebrate what we did with the last year.

WILLIE: Yeah, and I think also what is very important for us is to equip students and academics from all over Africa to come together, to meet each other, to network, but very importantly also to go back to their own communities and carry on this message of strengthening machine learning and, yeah, really just catching up with the rest of the world. Africa has enormous potential, and I think it's a very exciting time for all of us.

MELANIE: Agreed. Well, as I mentioned, we have some interviews in this episode. In particular, we are going to be talking with-- and Willie, you helped me with this-- we're talking with Yasin, and we're also talking with Vukosi. And we're going to spend part of that interview, really, talking a little bit more about what Deep Learning Indaba is about and some of the awards that were given out. And then the other part of our interview that was covered was with Andrea, who's VC here in Africa. And I know Nyalleng and I just spoke with her and had a great conversation.

So before we get into our interviews, as always, we start out with some of the cool things of the week and a question of the week. And we're going to wrap up with a question that is, pretty much, what are some of our key takeaways from the conference? But as I mentioned, we'd like to start out with the cool thing of the week. And we're going to do a little variation off of this, which is to help talk a little bit about what are some of the highlights that have been going on-- not the key takeaways, but things that have been happening during this conference? What are some of the great things that have been going on, or activities? Like I know that we've had a lot of speakers here, and--

WILLIE: Yeah, so we've really had an action-packed week-- great speakers. Nanda de Freitas was here, and Moustapha Cisse, and lots of others-- David Silver, Jeff Deane. So it's a really, really great list of speakers. We also had poster sessions, which were amazing, all the students presenting their own work, movie screenings of "AlphaGo."



WILLIE: There was a women in machine learning event one evening, which was really inspiring.

MELANIE: Nyalleng, you were speaking at that one as well.

NYALLENG: Yeah, and it was very well-attended. It was one of our biggest sessions.

MELANIE: And then there's been students who are from all over different countries within Africa-- my understanding is like 30 countries. And the number of attendees is around 550 attendees.

WILLIE: That's right.

MELANIE: Outside of the WIML event, I think one of the cool things that I saw was when David Silver walked in. And somebody was sitting there and looked over and saw him and was like, oh, my god, I just saw you in a movie.


He was so excited. I thought that was pretty fun to watch.

Any other little highlights that come to mind?

NYALLENG: Yeah. I Think I also saw somebody on a David Silver spotting. He kind of lost his mind a little bit.

The cool events-- oh, goodness. They are just so many. My coolest event was the award sessions. And the posters are always really rewarding and enlightening. And half the time, I mean, you just go there to just learn.

Meeting all the other people that are actually machine learning-- and maybe that's actually something to mention about the number of people that have traveled from far-- other researchers that have jam-packed schedule that traveled from far to come spend the week with us. And that's been very exciting to see them also.

MELANIE: I agree, very much so. Any posters that jump out?

NYALLENG: Oh, well, this is really putting me on the spot because the one poster that I do remember was actually Steve Krone and Herrmann. Herrmann is also a colleague who organized this. And this poster will be at NIPS-- this year's NIPS poster-- and they have a closure. They have discovered about how to add to noise in a network. So how much noise can you add in a network until you can't learn anymore? So actually, you saturate the network, and you cannot anymore learn in a deep net.

And so they have these closure properties about how deep you can go and how much noise you can add. And that's really beautiful, because so much-- I mean, that just narrows the parameter tuning such space. I mean, this is work that's being done right here in Stellenbosch.

MELANIE: That's great. Well, I think it's time for us to go ahead and jump into our interview, so let's go do that.

Today, I'm excited to have with me Dr. Vukosi Marivate and Yasin Musa Ayami, who are both here at Deep Learning Indaba, which is based in Stellenbosch, South Africa, for this edition of the Deep Learning Indaba. And with me also is Willie Brink, who is helping to be my special guest co-host. So thank you all, for coming today.


VUKOSI: Thank you

MELANIE: Now, let me go ahead and kick us off like we usually do, which is to have you tell us a little bit about who you are. So Vukosi, can you start?

VUKOSI: Thank you for having me over. My background, mostly, is in machine learning. At the moment, I am the chair of data science and a senior lecturer at the University of Pretoria. Taking you through my background, I was interested in machine learning and AI from a young age. I ended up going to University of Witwaterstrand in Johannesburg, doing electrical engineering there in undergrad and then a master's degree. From there, I moved on to do a PhD in computer science with a focus on enforcement learning in New Jersey, at Rutgers University.


VUKOSI: When that was kind of ending off, I decided that I wanted to go into data science because I was very interested in some of the-- kind of it's being dynamic and looking at different methods for lots of different areas. And then came back to South Africa, settled at the Council for Scientific and Industrial Research for about 3 and 1/2 years or so, working in data science. And that's how I also got kind of roped in much into the deep learning--

MELANIE: You got roped in. They found you, and then they made you come and do this.


Well, good to know. And Yasin, can you tell us a little bit about your background?

YASIN: My name is Yasin Musa Ayami. My background is in software engineering. I did my undergrad at the Durban University of Technology. And I graduated with a BTech cum laude. And then later on, I kind of transitioned to the field of AI.

My area of research was computer vision. So besides that, I'm also a certified Oracle associate. Also, I'm the co-founder and CTO at TsogoloTech. And yeah, my vision has always been to kind of leverage tech for social good.

MELANIE: Nice. That's great. Well, so the reason why we're doing this recording specifically is because we wanted to touch on two things that are playing out-- one, Deep Learning Indaba, and what it's about, as well as the Maathai Award that is at Deep Learning Indaba, which Yasin, I know you had won. We were actually just talking about it today.

So let me go ahead and kick off with, first, what is Deep Learning Indaba? Tell us a little bit about what is this conference? Where did it come from? What do you know about it? Why are you doing this, Vukosi-- and Billy?


VUKOSI: I'm being facetious about being roped in. It was a good group of friends and acquaintances who came together after one-- I guess, a number of iterations of-- if you went into some of the big machine learning and data science or data mining conferences, seeing the lack of participation from the African continent is something that was, I think, concerning. And not more [INAUDIBLE] do people have papers there, really, but on-- I think, from my view, that those people are missing out.

There's a lot that's going on the continent. There's a lot of very interesting areas that people are working on that might not be seen-- maybe if you're sitting in a Google or a Microsoft in Europe or in the US, that you might not think about, but then people are trying to resolve.

So through a number of conversations, I knew a lot of the co-organizers personally before. And we started talking about this and saying, can't we do something in terms of trying to grow machine learning and its visibility on the continent? Not that it's not there, but if we could find a way to bring people together, come share their work, their expertise, build a network, and get this to kind of snowball into something that's going to be much bigger than just us?

So the Indaba really encompasses that, in that we're trying to strengthen African machine learning. And by doing that, it's building the community in itself, as one way of doing that-- learning from each other, offering each other's skills to the community, and seeing where this is going to get us in a couple of years. So even though this is the second year, it is still like the beginning.

MELANIE: And you mentioned it's the second year, last year being the first year. Where was last year? And how many people?

VUKOSI: Last year was at the University of Witwaterstrand, or Wit University, as it's known in Johannesburg. And there was about 330 people from 22 different African countries.

MELANIE: Nice. And this year, it's around 550 is my understanding.

VUKOSI: 550, yeah, as a target.

MELANIE: Nice. Well, when you mention community, I know we were talking, actually, before we started recording about how there's all these Deep Learning Indaba X events and community-building activities that have been happening throughout different countries within Africa. And so Yasin, I know you are one of the organizers. And specifically, you organized one in Zambia. And when was that that you organized it?

YASIN: On the 24th of March.

MELANIE: OK. And what was that like in terms of organizing the event there?

YASIN: It was a bit challenging, organizing the Indaba X there, because one was the issue of resources. Having spent most of my time in South Africa in organizing things here, I was assuming that it's going to be the same [INAUDIBLE]. But when I got there, I literally had to pay for everything, including venues, computers, and things like that.

So when we advertised for people to register for the Indaba X, we received a number of applications. But sadly, because due to the number of resources, we could only accommodate a few. And we only accommodated about 60 participants. And other participants, unfortunately, we could not accommodate them.

And mostly, if you look at the expenses that were incurred, you know, there's expenses that we knew were supposed to be incurred anyway, because you're looking at something that-- you want to empower people. You want to give knowledge to people. You want to produce a cohort of youths that are technologically empowered. So we're trying to associate it to universities.

But when we approached universities, they were not open to the idea. You know, because they were looking at it from a marketing perspective. I mean, they were talking about issues to do with, OK, if you want to partner with us, come with at least 50 participants. So that, in a way, was difficult from our end, because we couldn't provide for the 150 people. So we ended up resorting to having other venues, which was not supposed to be the case.

MELANIE: Well, what went well for the event?

YASIN: OK, so what went well is the positive feedback we got from people. So people now are saying, let's make this an annual thing, whether Deep Learning Indaba is in, Deep Learning Indaba is not in. So we've got also some good responses from corporate people. And we're hoping that more would buy into the idea.

So we've had people coming to us and saying, OK, why you never reached out to us? And we're like, no, the people that you've employed to let us have access to you were denying us access to you. Or sometimes we'd send emails, but you were not responding to our emails. But at least, now, we've kind of established contact with these people, and we are able to contact them directly.

So going forward, I think the subsequent Indabas will be bigger and better. And obviously, we'll have more and more people coming on board. So the next Indaba, we're hoping to have at least maybe 150 or more people. And yeah.

So arising from the Deep Learning Indaba, there are some monthly tutorials that aim at increasing the number of people that are in this area of data science. And yeah, so we're hoping that as the population grows of people who are in data science, then certain issues-- you can have a number of people working on various issues that are being affected in a country.

MELANIE: What's going well? And what are some of the challenges that you've seen, Vukosi, from working on Deep Learning Indaba?

VUKOSI: The biggest challenge is always demand.

MELANIE: Very common to what Yasin's saying, yeah.

VUKOSI: Yeah. Starting off, I think, maybe going back to the beginning of the Indaba, we had envisioned, I think, 50, even in the first year.


VUKOSI: That was what we were planning for. And we ended up getting like 700 applications. And by then, I think we were thinking, OK, let's just make it 200. So by the time 700 applications came in, it was, we can't do 200. We need to increase this as much as possible.

So it's something that we have to come to terms with. And again, it's the thing of saying, things are happening. People are wanting to do more in this area.

The amount of times you talk to professional students, even academics, who come up and say, look, yeah, I'm doing Y or X, but I'm actually very interested in machine learning and artificial intelligence because I have this idea. And I think if I can just grasp the methods, I will be able to actually do something with it. That's what the Indaba is for.

And the knock-ons are still something that I think we're still having a very hard time being able to quantify. So just from what Yasini is saying, of saying, now, you have monthly events. And every time you meet the Indaba X organizers, there is always a story about there's this other thing that we actually started doing afterwards. And being able to empower them to continue into their journeys is also important, because we, as the main Indaba organizers, can only do so much. You can only cover so many.

So even with the first year only having 330 being at the main event in Johannesburg, the Indaba X's actually then had a total over 1,000 over that March and April period. So you can see that that's a bigger factor in terms of doing that. And yes, it does bring in more people, more demand, again, to us, because then people want to come for the--

MELANIE: But that's a good problem, right?

VUKOSI: Yeah, that is a good problem on there. And yes, for us, we try to keep the mission being the main thing, the strengthening of African machine learning, as being what we want people to sign on for. So in working to get sponsors or get partners, maybe [INAUDIBLE], we want them to buy into that mission. And if it's not the thing that they want to do, we understand. But it's this cornerstone that we need to keep, because that's why people come here. They sit over six days and with little sleep going through a lot of content, talking to each other. You get amazed at some of the connections that are made.

MELANIE: A significant amount of content that's definitely being covered at this conference. So I can understand that. And the fact that there's monthly tutorials that are also being produced is also kind of incredible. And frankly, I heard someone saying how they were amazed at how much people are just engaged. It's so wonderful to see how much engagement there is, how much excitement there is, about what's being shared. So thank you.

Well, OK, so let's get into the fact that there's two awards that were given out, and some honorable mentions in relation to those awards. And I know the Maathai Award in particular was given out. And it was really to recognize along the lines of even people who are individuals especially, who can have large impacts in the world around them, especially in the ways that they approach problems that they're solving. Is that doing it pseudo-justice? I know it's probably not doing it full justice, but could you expand a little bit more on that, Vukosi?

VUKOSI: It was a prize to recognize innovation using machine learning or AI across the continent. So whether it was a group or an individual, it didn't matter, but we wanted to see what people were doing out there. I think Yasini completely embodies this, where you go off and you have this societal challenge that is out there. And you think about, what is an approach that I could use to try and tackle this? Right?

And then you go backwards, and then you find something that's kind of tangible. And then you say, OK, I need to now use this tool and learn and see how I can innovate around it to be able to then get people where they are, as opposed to going, hey, I've got this hammer. I need to hit nails. And I'm looking out for nails. And I think, yeah, that's what the Maathai Award is about.

MELANIE: So let's talk about what you're working on-- TsogoloTech.

YASIN: TsogoloTech, OK. So TsogoloTech is a tech startup that aims at making use of state-of-the-art technology to contribute toward solving these socioeconomic problems that we're currently facing. So among them is the issues of cholera, the issues to do with financial inclusion, like underserved communities. How can we include them in the formal financial sector? What made me win the Maathai Impact Award was the app that we created, called Tutuka app, which aims at automating transactions.

What happens is women in underdeveloped areas have resorted to coming up with village banks. So village banks, this is kind of a system which enables people that are living in the same area to contribute money. Then from that money that they contribute, then they are able to lend it out amongst themselves. And then they're able to repay back interest.

So what was happening is-- of course, that was a good idea, but the challenge was the transactions that occurred during this process were done on paper. So we kind of thought that, if we automate this process, and once we automate this process, then we'll have data, data we can use for many things.

Among things that can be done with this data is approach banks. Because you find that these areas maybe do not have a bank. So you can go to your bank and then say, OK, let's see, these people are making X amount of savings. And I think these people need a branch or something like that.

Or again, since there's this issue of cryptocurrency, the blockchain that's making waves on the internet-- I mean, how can we innovators come up with ways in which we can use these technologies to aid these people in managing the transactions? So that's what this app is all about. And we envisage, by 2019, to reach out to about 22,000 women.

MELANIE: That's awesome. Well, why women?

YASIN: OK. Well, if you look at the stats, financial inclusion stats in Zambia, about 41% of adults are financially excluded, and majority of which are women. And most of these are rural dwellers. So there's a reason why we're mostly looking at these people, because these are people who may be going door to door selling items, farming. So the little that they are making, if it's not properly managed, then they end up having nothing. So I think that's what made us focus mainly on women.

MELANIE: That's wonderful. And the app itself, can you tell us a little bit about that and how it functions? What kind of technology are you using?

YASIN: OK, so we had created a prototype, a mobile app.

MELANIE: It's called Tutuku?

YASIN: Tutuka.

MELANIE: Tutuka.

YASIN: Tutuka, yes, which enabled these people to enter their transactions on this app. And then, obviously, there was some back end application that was storing the data.

So now what we observed is the literacy levels are very low. Also, the financial literacy levels are also low. So you find that there were issues in terms of recording these transactions. And you find that somebody cannot write or record their transactions, so they had maybe to depend on somebody to record for them transactions. Consequently, somebody might end up maybe recording wrong transactions, or one is not able to track their own transactions.

So if we automate this process, we thought this could somehow knock out some of these issues. So that's the reason why we created that app. But since these people have low literacy levels, it's been difficult to make use of the app. In looking at the penetration of internet levels, it's still not that much. So we're thinking of now integrating the SMS system, so where we want to use USSD for them to record these transactions, instead of using the mobile app.

Then subsequently, after we've got these transactions automated, we're looking at making use of state-of-the-art machine learning, deep learning, to kind of help these people build their credit profile. Because if you go to a bank, obviously, they will need some form of collateral for you to get a loan. And now these are people that do not have access to banks.


YASIN: And of course, many, they don't have access to loans. Just as an example-- I don't know if you've been following the news-- we've had, three infernos in the last couple of years, which happened, I think, two months ago-- three months, COMESA Market, which houses people from the COMESA region. So you've got people from Tanzania, people from Malawi. You've got people from Zimbabwe.

So there was an inferno there, and people lost goods worth--

MELANIE: But they lost their currency.

YASIN: Yes. So you find that that was their only source of livelihood. And they couldn't be compensated. They were looking up to the government to kind of compensate them, but the government couldn't. And so these people have just resorted to staying at home.

So Zambia being a third-world country that is striving to fight poverty, I don't think it's possible to fight poverty, if that's the case. So we're hoping that, if the app is successful, then we can help them build their credit profile and then subsequently link them maybe to formal financial institutions, like the banks, where they may be able to access loans.

Because, of course, with village banks, they're able to get loans. But it's only a limited amount of money, because the contributions, it's maybe $1.00 or $2.00. So if later somebody wants a loan of maybe $1,000, so you find that maybe the group cannot sustain that. The amount of savings that they've got cannot service that. So if we build the credit profile, then subsequently, these people will be able to approach the banks. And then the banks will be able maybe to give these people loans. So that's what we're aiming at.

And obviously, there's been a lot of research that's been going on. I mean, attending an Indaba, we're talking about 550, and majority of which are researchers. So if you carry out a random survey, I don't think there's maybe anybody that's maybe working in the financial inclusion space, or to be specific, somebody working on this village app thing.

One of the reasons could be the issue of data. So we're hoping that if we generate this data, we can then push it to the public so people can make use of it and then do some research. And of course, we are cognizant of the fact that, yes, there are issues of ethics that have to be considered. So that's where we're heading toward.

MELANIE: So in terms of the data itself, you said you're storing it currently. What are you storing it on?

YASIN: Well, we're just storing it online. We've got a back end application that's storing the data. And of course, we are going to use some APIs to expose the data, if we want to use it for machine learning or something like that.

MELANIE: Is it an app that exists currently?

YASIN: It's a prototype.

MELANIE: Prototype, ah.

YASIN: But it's not yet on the market. I mean, we're just testing out.

MELANIE: Got it. What are you prototyping it on? Is it on an iOS or an Android?

YASIN: Yes, it was an iOS app. Used some Microsoft tools. I don't know if I'm allowed to kind of market other people's products.

MELANIE: That's OK. Yes, you can totally talk about what other things you use. What are some of the ones you used on the Microsoft side?

YASIN: I was using Submarine. Submarine enables you to create cross-platform app. OK, so with TsogoloTech, we don't necessarily have in-house developers. So mostly, it's people who've got time and are willing to spend their time working on social issues.

So it was one of my friends, Janine, that was actually working on the app. We were actually collaborating. I was doing minor changes here and there, but she did most of the work. And with me, I was focused on creating the back end and, obviously, providing her with the necessary API's that would enable her to consume for the app functionality.

MELANIE: Nice. What are your future plans?

YASIN: My future plans are to tackle as many problems as possible. Of course, offline, we were talking about the issue of cholera. In the next three months, there will be another disaster, I presume, because there's nothing that has been done from the last time we encountered this problem-- the issue of cholera.

So I'm trying to also see if we can tap into that also, to see how we can use the existing data to reduce the number of cases that are recorded, and subsequently reduce the numbers of deaths that are recorded. So those are some of the plans. And yeah, also, to get as many local developers as possible. Because I don't know, I don't see it appropriate for local developers to contribute to offshore projects, or in GitHub, when we've got a lot of issues that have to be tackled in-house.

At Tsogolo, we are fortunate to have two groups of people. One are engineers, and the others are people that are working in developmental work. So these people working in development will kind of point us to areas which need attention. They tell us, OK, I think this needs a problem. So we engage the engineers to come up with a tech solution that can tackle this problem. So we're hoping to have as many local developers as possible helping towards achieving this goal.

MELANIE: Do you accept support from external developers that are not-- you know, is that something that you do currently, or you want to do?

YASIN: Well, I think it's something that, yeah, we'd want to do-- collaborate with certain developers. I mean, on Saturday, I'm actually leaving. I told you I'm leaving for Germany. So with Germany, we're going for a certain program that's more like an incubation program. So they walk you through the stages of developing a product and, of course, releasing it to the public and, of course, ways you can manage that product. So it's basically aimed at startups and then seeing how they can sustain this startup. Because what happens is you've got a startup that is in existence for a few months to years, then it just disappears.

MELANIE: Right. Well, that's great about the startup incubator in Germany. And then, in terms of if somebody is listening to this, and they want to help, or they want to participate in terms of what you're building, how would they do that?

YASIN: So they can reach out to us.

MELANIE: We'll definitely have your contact information. We'll get that from you and put that on our show notes. Yes.

YASIN: They can reach out to us. Because the issue is we've not been able to do much, because we just are relying on people that are already working. So they can only work on certain projects when they are free. If we could have proper funding to have in-house data scientists, software engineers, we will be able to work on these problems more quickly, you know?

Because the issue with tech is-- like they say, the only thing with technology that does not change is that it's always changing. Maybe today, I might be talking about, OK, we want to make use of cryptocurrencies and blockchain to tackle this issue of financial inclusion. Tomorrow, there's something else. So once I'm still working on this, then it becomes, in a way, obsolete. So I just have to stop, because I feel de-motivated. Our goal is to come up with ways in which we can quickly work on these problems, and then maybe deploy it to the public, and then see if challenges that people are facing, analyzing whether this problem is solving the problem that we're trying to solve, then also trying to identify ways in which we can improve on the product.

MELANIE: That sounds great. Well, thank you. Before we wrap up, what I want to touch on is, what are some of the key things that you're taking away from the Deep Learning Indaba for this week? What are some things that you've experienced so far that have been really valuable for you, in terms of what you've seen?

VUKOSI: Just sort of touching on what Yasini has been talking about-- my agenda now at University of Pretoria starting out, especially in data science, is to do data science that comes from a social bent, so coming from social impact as a space. So I'll be very happy to have more conversations about how do we go from that data, find ways to release it in the proper way for researchers to be able to use, and for us, at a place like the University of Pretoria, be able to work with students to actually build on the research in that space. Because yes, on one part, there is data. The other is experts, like Yasini over there.

And this also has permeated the Indaba. If you go in and you see the posters, and you see what people are actually presenting and the things that they're working on, yeah, you start seeing threads of, oh, this could turn into something bigger, if you could get one, two, three, four people together from these different countries. So we shouldn't look at our borders and say, that's where we end. I think the Indaba is trying to reduce that distance and allow people to be able to work across and be able to mentor students, mentor professionals across these borders.

So I'm hoping that is a thing that is going to go on. I've seen examples of people doing this after meeting, especially another meeting-- not your, like saying, advertising. At Data Science Africa, I did meet people last year as a part of my trying to cover as much of the continent as possible and bring people in, just like we met [INAUDIBLE].

MELANIE: We did. I know. And I was saying this offline, that I know about this conference because of you, Vukosi, because you were at a conference that I was at, and you spoke about it. And I immediately was like, how do I get involved?

VUKOSI: So, exactly. I was at Data Science Africa trying to talk about the story, like show about my research, but also that we're having the Indaba. That led to-- we have one of our co-organizers, Kathleen, who actually I met at Data Science Africa. And I had a gentleman that I got to then mentor this year, working on disease modeling in Malawi, just having a few Skype meetings. It's not this-- like people think, oh, mentoring means you're spending half your--

MELANIE: Years, and hours, and all day. Yeah.

VUKOSI: Yeah, yeah, and hours, and doing that, but then trying to reduce that and being able to-- and I've seen people like Benjamin Rothman do the same thing.

WILLIE: Yeah. And maybe I can just add that Africa, as a continent, has so many problems that are unique to the continent. And it is very important for us to equip our students to be able to deal with these problems. And hopefully, the Indaba also goes somewhere towards that goal.

MELANIE: That's great. And I know that the sessions that are being held right now are being recorded. And that will be released. Do you guys know when that's going to be released?

WILLIE: We're hoping for very soon after the Indaba, maybe over the weekend or early next week.

MELANIE: Great. And-- oh, and you were going to say something else?

VUKOSI: It also again speaks to this ethics of sharing. Whether somebody is at the Indaba or not, should get as much of the content as possible. So there's a lot of work that has gone into the practical sessions. And all of that is on GitHub already. And people can go interact with it.

MELANIE: We'll get the links and put those up there.

VUKOSI: I keep on saying, we're trying to manufacture abundance. It's not about reducing access, right? So it's increasing it, so making sure that there's videos, there's the practical sessions. And the practical sessions, also, all that content is available.

Yes. Sometimes, it does go down. When it goes down, it's because it's being improved. It's not that we're removing it from-- and that's what happened. You know? We switched off last year's practicals. You then start getting messages with people asking you, please, can we have access? We were actually just waiting for the switchover to this year's, which is practicals which are much better.

MELANIE: Yeah. I've heard wonderful commentary and people seeing the practicals and having wonderful experiences with them. And next year, Deep Learning Indaba is going to be in Kenya. Do you know where?

YASIN: Kenyatta University in Nairobi, Kenya.


WILLIE: The good news is also that we are again trying to increase the numbers quite substantially, so please look out for the call for applications for next year's edition.

MELANIE: Great. Anything else that you wanted to talk about or make sure we cover before we go?

YASIN: So I just wanted to add to on what Willie and Vukosi were talking about-- issues to do with the problems that we're facing as a continent. So one problem that we're facing, as Africans, is when we get people like Google, when they approach us, so we leave Africa and then go stay in Europe. In terms of making progress, we are, in a way, retrogressing. So we're not making any changes to whatever problems that we're facing.

So I think, of late, yes, we've had issues where you've had to hire experts from Europe to come and solve our problems here in Africa. And of course, I don't know, for lack of a better term, I would say, an oxymoron to say that, OK, we've got-- there's unemployment in Africa. On the other hand, you're outsourcing people to work for your people.

So I think right now, since AI is still in its infancy, if we can get as much people as possible into this space of AI, so that instead of outsourcing people from Europe to come and solve our issues, I think we should be the better people to attack all our issues. I think we understand them better-- like we all know you cannot compare an original version to a translated version.

So I'd like to recommend [INAUDIBLE] that these guys, like Shakir and other guys that are working in London, who-- of course they've got jobs there, but they did not forget about this African continent and, of course, decided to come up with this Deep Learning Indaba and, of course, sacrificing their time, their efforts, in trying to make this thing a success.

From these 550 participants, you can just imagine how many more people are going to benefit. Obviously, people are going back to their communities with feedback. And of course, the Deep Learning Indaba X's, I think it was in 13 countries, the previous ones, so we've got about 22 African countries. So I'm hoping that, in 22 African countries-plus, there will be some sort of an Indaba X. And then if the Indaba X is usually held around March, if we can have total participants for these Indaba X, like even more than 100,000, so I think it would could be a great plus to me.

So we need to just say that, OK, they're trying to increase the number of people that are going to participate in the Deep Learning Indaba. So people don't necessarily need to attend the Deep Learning Indaba-- the main one. But through these small Deep Learning Indaba X's-- because of course there is the issues of funding. So you have to fly people from Cameroon, people from Egypt, to come and attend in Indaba. So that, in a way, would limit you in terms of doing certain things. So if we can identify people that we can train properly and, of course, run these Indabas in respective countries, I think that would be more effective.

So as I was organizing the Deep Learning Indaba X, one of the issues that I faced was the issues to do with speakers. There were few speakers I knew of that could deliver talks at the Deep Learning Indaba X, one of which is Cynthia. She's here. She was invited to come and to do sessions-- Maps tutorials.

So if we can also boost capacity in that area to increase the number of people that can also offload this content to other people, I think also would be heading towards the right direction. So I don't know if we can have Indaba just specifically designed to train people to come-- sort of a trainer of trainers, which-- yes, you know, when we talk Indaba X Zambia, we know exactly that, OK, there's X, Y, and Z. Because one is having an Indaba X in Zambia, but two is also having people who are able to deliver content to these participants and then, of course, having these participant's benefit from the tutorials that will be given by these participants.

MELANIE: This is great. And thank you. And anything else that you guys wanted to touch on before we go?

YASIN: The thing I've been parroting over the last few days of just talking over and over again is going to be that, yes, we are getting to a point where we have to think about how do we grow? Scaling definitely on our minds, as organizers, and trying to figure out what's the best way.

At the same time, we aren't ignorant of the fact of a lot of other initiatives that have been kicked off, or had kicked off a bit earlier than us, or have kicked off after we had started. Even here in South Africa, there was a lot going on in this week in this area in terms of machine learning. So just think about the continent. And so, how do we come together as this wider community?

MELANIE: And be stronger.

VUKOSI: And then be stronger? Yes, it's not about us. That's a thing I'll keep on saying to my co-organizers, that it's not about us. It's going to be about the community. And to do this, we're going to have to come together and come up with a strategy on how we bring in all these resources that we have. They might be minor in the bigger scheme of things, but we still have to optimize them to deliver well. And through this, if we can connect them together, find ways for people to move in and around them, it builds a much better, a stronger community of practice.

Right now, there's International Data Week, that will be happening in Botswana in November. I am hosting a session there on Data for Good, specifically to try and find even more people who are working creating these honeypots of people who will submit some abstract. And then you'll go in and say, oh, we didn't even know that people in your area are doing this. So maybe you should also come in, and things. So that community-building still has to go on.

MELANIE: And let me also end on this point. If people are interested in getting involved, whether it's participating in the actual conference, whether it's providing resources, whatever it is, how do they get in touch? Who do they get in touch with?

VUKOSI: At the moment, I think, on the website, we've set up a link to try and get as much information about researchers, practitioners, professionals on the continent in machine learning, in AI and data science. So you can go and actually sign up and submit a form. We're trying to keep track. We have a mailing list that people can sign up for. They can send messages through that, in terms of the community that's connected to the Indaba and beyond. That's the ambition there.

And through the website, there is a contact. And if you send an email to that, it will then get in touch with us directly. All the organizers actually man the email, so somebody will respond to you quickly, in terms of what you can do. So the website is the lightning rod, and you will get all this information there.

MELANIE: Great. Vukosi, Yasin, Willie, thank you all for joining me today. I really appreciate it.

WILLIE: Thanks, Melanie.

VUKOSI: Thank you.

YASIN: Thank you.

MELANIE: Today, I'm very happy that Andrea Bohmert is here with us. And I have a special guest co-host with me, Nyalleng Moorosi, who-- I know she's looking at my notes, because I wrote down her name. And I didn't write down the right spelling, but I wrote it so I could read it. Anyways, thank you both for being here.

ANDREA: Well, thank you very much for having us.

MELANIE: You're here because we're at Deep Learning Indaba. And you are here to give us a perspective from the VC perspective, in Africa in particular. Can you tell us a little bit about you and your background?

ANDREA: Sure. Absolutely. So as you can hear from my accent, I'm not originally from South Africa. I'm from Germany. After finishing university there, I came to South Africa for an internship and kind of stayed. So that's now 20-something years ago. So I've spent the last-- or my professional life, actually, in South Africa, first few years working for a big corporate, and now since 12 years, actually, in the venture capital investment space here in South Africa.

MELANIE: Thank you. Well, so what has the Deep Learning Indaba conference been like for you so far?

ANDREA: This week is actually a very, very interesting week. I was unfortunately only able to be here for the first day-- for the opening day-- because at the same time, in Cape Town, we have three other conferences kind of happening. So I'm actually very, very excited to be here. I must say, so far, I've been very, very impressed with a few things. First of all, the sheer quantity of people, the whole diversity, the fact that there are people, I think, from 33 countries here. And very honestly, I was quite impressed by the number of females. Normally, one kind of thinks data science, that this is still a very, very heavily male-dominated kind of-- and I was very positively surprised. Quite a few women actually are here.

And I love the communication-- the level of communication. I thought that data scientists are introverts. Apparently, they are not. They actually talk.

MELANIE: We are introverts. [CHUCKLES]


MELANIE: Some of us are.

NYALLENG: Yeah, some. I actually think we might have a mix of some extroverts and introverts, but I think we work well together.

ANDREA: I thought everybody would just stand there looking down, being very quiet. And I walked around, and everybody was chatting. So I think that's amazing. That's great.

MELANIE: My understanding is that you are one of the biggest investors in AI in Africa. Is that statement correct?

ANDREA: Well, so let's put it like this. I think we are one of the oldest VC funds here. And yes, we've always done tech investments. And we have done an investment in a AI company.

I think it's quite interesting, because not many African investors have so far invested in AI or machine learning kind of companies. A number of international have. If I want to talk about how we're one of the very few African slash South African investors who have invested in an AI company, then yes, absolutely.

NYALLENG: You know, you mentioned something about the number of women that you've seen here. How has the number of the women in the VC, especially in your area of investing in AI in Africa? Are you just the one data point?



Yes. No, I think I'm not the one data point, but I think in-- and I can actually not really talk about the African, I have to refer to the South African. But I think in South Africa, there are maybe three, four female investment partners. Yes, it's a very heavily male-dominated world, and we are working on it.

MELANIE: From what you were saying, why is it that not as many VC's in the African continent are doing investments in AI in Africa?

ANDREA: I think, first of all, you should invest in things that you understand. And now before you ask me whether I understand AI, that's a different story. But I think a lot of people are a little bit scared, because AI is currently used in close to every pitch. There's a lot of hype about it.

There's buzzword bingo, absolutely, around. If I look at the business plans that come across my table, I would say every second has got a word like AI or ML somewhere in the business plan, which doesn't mean that they're really doing it. And I think a lot of investors are scared to fall for the hype because the moment, of course, a company is really an AI company, they ask for much higher valuations.

So that's why a lot of businesses are now trying to go and sneak those words in. So they employ one data scientist, who maybe actually just does a bit of very advanced Excel modeling, and they call it now, (IN REVERENT, HUSHED TONE) they're doing AI.

And I think a lot of investors are still quite nervous to understand, is this now a company that really knows what they're talking about? Is it now really a company that applies deep algorithms and understands what they're doing, or is it just one of those hype companies? And I think that's one of the big reasons for it.

MELANIE: Well, what does AI mean to you, as an investor?

ANDREA: For me, it means that you're not just slicing and dicing existing data, but actually that you can go and start using this data, teach it to go and get information out of itself, and just taking it to the next level to come up with insights that were previously not there. And this is far more than just slicing and dicing the data, it is taking it really to the next level. And for that, you need to have some very, very ingrained understanding and skills. And a lot of companies simply do not have them. But they use the buzzword without having the capabilities and without actually applying it in the way that it should happen. I think that's where one of the big, big differences are.

NYALLENG: Can you just walk us a little bit on the history of your company and how you ended up here in this fairly unique space?

ANDREA: In terms of in South Africa, or at the Learning Indaba, or in terms of--

NYALLENG: Oh, no, not the Deep Learning Indaba-- as a VC.


NYALLENG: In the history of your VC company.

ANDREA: Sure. How I got into VC is actually quite a funny story, but it might be kind of too long. I just wanted to be a VC. And there was no venture capital, so I managed to go and raise a fund from a high-net-worth individual, the previous founder of SEP. And that's how I got into VC.

Our fund itself is now about 12 years old. And it's quite interesting. In South Africa, we've got a very, very established-- oh, actually, in Africa, private equity industry, everybody wants to invest in infrastructure. But when it comes to investing in technology, there's actually very little. A lot of people still say technology is too risky, early-stage technology is too risky, and early-stage technology in Africa is definitely far too risky.

So therefore, if you're going to go and look at it, there is not much institutional money, like in the US or in the world. The VC is actually there founded by institutions. In Africa, most of them are actually founded by high-net-worth individuals who are passionate about this and have put certain money into it. The problem is that most of those funds are relatively small. So if you have a $20 million fund in Africa-- South Africa-- this is a huge fund. Internationally, of course, that means nothing.

So one of the things is really how to go and start making the right investments that don't need $10 million, $15 million, but actually need $2 million, $3 million, and really go and take them. And we, as a fund, have done this now for the last few years. We have just closed our first fund where we had seven investments. Out of the seven investments, we had four exits to global, Fortune 500 companies. We are one of the few funds who can really go and say that we have closed the whole investment cycle.

Investing? Everybody can invest. The question is, can you make returns out of it? And can you make returns out of it that not just make you as an investor happy, but make actually the entrepreneurs and their teams happy as well? Because it is a long-term journey. In venture capital investment, you can't be short-term greedy. You have to be long-term greedy. It is a long, long time when you have to go and work together with companies.

And yeah, we've done this where we've closed our first fund very, very successfully and at a high return. And now, we are busy investing the second fund. And yeah, I am very, very excited to go and see how AI and ML and how companies who are really solving African problems using this kind of new insight and new knowledge that actually exists, how we can go make some cool investments in that space.

MELANIE: A couple of different questions are coming to my mind right now. But you know, one of the things in particular that I know we talked about briefly before was, how does this compare-- startups that you're investing in-- compare to those that you've seen out of Silicon Valley, for example?


MELANIE: What does that look like?

ANDREA: So if you think about it, as I just mentioned, our funds are relatively small. So if I can only go and put $2 million or $3 million into a company, the Silicon Valley counterparts, they get $10 million, $15 million. Sure, we have a slight cost advantage here in South Africa, but that only applies while we are busy here in South Africa. So the moment we want to go into business international, we still-- we are traveling. Everything still becomes US dollar-based and so on.

So I think the big point is, because our funds are smaller-- therefore, the amount per company are smaller-- our companies have to become real businesses. That's an advantage and a disadvantage. So we cannot go and take a company that can go and be not profitable for the next five years, who can just go and employ, employ, employ people, maybe just go and be very, very academic and come up with amazing results, but don't actually have a real underlying business.

In Africa-- in South Africa-- our businesses have to become, very, very quickly, businesses that are making a profit because there is not enough money that you can go and pump into them all the time. So therefore, there's a slightly different mindset.

If we compare the companies that we invest in to their competitors in the US, for example, then we quite often see that they have sometimes even bigger teams. They don't have case studies. They don't work on return of investment. They do a lot of consulting work, but actually are not really worried about, I'll be profitable. I'll be making money. Because there's not this kind of pressure on them, because their investors will just fund it and fund it.

In our case, we tell them, sorry, you've got $2 million, $3 million, you have to make it work. At the end of that, you need to be profitable. So our businesses really-- their focus is right-- they have to employ sales people. They have to employ marketing people. It's not just about the science, it's actually about building a real business.

MELANIE: That makes sense.

NYALLENG: Mm-hm. I think one of the things that is always raised, especially when you get to the space of entrepreneurship and obviously funding entrepreneurs in Africa, is that maybe the market or the environment is not ready. I can see you maybe nod, yes. This is exactly what I would like to hear is your opinion of, what do you think is there in the environment that either has enabled you and your company to be able to capitalize? Or what do you think is there in the market that actually might inhibit somebody else coming in with their VC and investing in Africa?

ANDREA: So I think that all the various ingredients of the ecosystem that is needed for successful entrepreneurship-- the ingredients are there, but they're not linked. They are still working in isolations. In some areas, in some ecosystems, it's working closer together. In others, it's definitely not, which makes it far more difficult. If you, as an investor, look to go and make investment, if you just wait for the perfect business plan to arrive in your email inbox, it's not going to happen.

So you need to be out there. You need to go and understand that you will find a rough diamond, and you know you have to polish it. The perfect diamond is not there. So you need to know how to identify rough diamonds, and you need to understand how to polish. That's the first thing.

From an entrepreneur point of view, I think, maybe because we don't have enough highly visible success stories out there, there's still not enough of the real amazing potential entrepreneurs that are actually deciding to become entrepreneurs. It's still quite tempting to actually join some of the big corporates who are, of course, paying much, much better salaries. And well, then life is far more secure. And it's not that obvious of how to really go and do it.

I think our academic institutions and even our schools are also not preparing students early enough to also think business. I think we have a lot of great specialists. But if you want to be an entrepreneur, you also need to be a generalist. You need to understand how to manage people. You need to understand a bit of finance. You need to understand a bit of sales and marketing. So if you are purely a specialist in a very narrow area, it's quite difficult to go and run a business where you have to be a generalist as well.

And it's also trying to go and find the right diverse team, right from the beginning. If you have a group of 10 data scientists, they might come up with the most amazing solutions, but I'm not sure whether they'll come up with an amazing business. And I think, so we need to go and start learning that it's different skills that are required.

How do we, as a country, as a continent, make sure that entrepreneurs actually have either access to those skills, or have been taught those kind of skills. And then all the other areas of the ecosystem-- the rules, the regulations, the access to funding, all those kinds of things, access to expertise, whether it's now legal accounting, all those kind of things-- all those things need to go and fit together. And then, I think, for the rest, I can definitely say that the entrepreneurs that I see are absolutely on par with what I see globally. It's the rest that needs to go and fit.

MELANIE: That was actually going to be my next question to you. Is that based on what you've seen so far in the different countries that you've been to within the continent? What's going right? What's working that's making it a good place to be investing in AI?

ANDREA: The passion is there. The skills are there. Seed funding is often missing, the kind of how to go and get somebody to take this leap to actually go and start it. Not everybody can afford to not earn a salary for a year's time. So how can we go ahead and close this gap? Access to market and understanding how to sell. I mean, if your potential client base is a big corporate, well then, a lot of people, particularly when they come straight from university, they don't know how to do this. They don't understand how to go and commercialize the business.

So where I see the biggest challenge is not when it comes to, could they develop an amazing solution? People here have a deep understanding of the problems that they want to solve. They can develop the solution. What's quite often missing is, how do I take this solution to market? How do I go and pitch to a big corporate? How do I nego-- what is the right kind of pricing for it?

So the challenge is really in the, how do I go and take a fantastic solution and build a fantastic business out of this? And I think we need to support. We need to go and come up with support structures, because that is where things are missing.

And if I then go to the US, for example, somehow, I think that gets far more ingrained from early on. There's different kind of support structures there. So there, it just seems to be a bit more easier. Maybe also because, there, the customer base is far more used to actually going to speak to startups. Whereas here, our typical corporates are still quite reluctant. It's the typical, nobody ever gets fired for buying dot, dot, dot.

I mean, people here rather kind of go and do business with large businesses than to kind of go and give a startup a real chance. And so I think it's those kind of things that quite often make a difference.

NYALLENG: I have two questions, actually, that just follow up from there. And the first question is, what drives you to do this work? I can tell already that you're speaking with a lot of passion, but what drives you to do it and to do it here? And then the second one is, what are the specific examples that are getting you very excited to be investing in this space?

ANDREA: I come from Germany. I could go back to Germany any day and start doing it there. But what I love about this and what drives me is the fact that there are amazing challenges here, but you can solve them. And you can make an impact, and you can see the impact. It's in so many different things, and I just love this.

I love to work with entrepreneurs. And sometimes, one phone call can make the difference. By helping with the one little insight, it can make the difference. It'll make a difference not just to one entrepreneur, but to the whole team, to everything. So I love that part.

I love the fact that, here, you can still challenge the status quo. You can break new ground. And I love the fact that success can stimulate so many people. And here, I see so many young people who really want to go and make a difference. And they could also-- with the right skills, they could go wherever. But they choose to stay here and build amazing businesses from Africa, to go and make a difference, to be role models. And I think that's what the country needs. And if I can play a tiny little role there, I think that's amazing. That's the fun part.

And I mean, I have children myself. Yeah, I think, if you want to go and prepare the next future for the world that's lying ahead of us, well, you need to go and invest in passionate entrepreneurs. That's just what I believe.

MELANIE: Any resources you'd recommend to those who are trying to start their own business, exploring the space?

ANDREA: I think one of the tools that a lot of people absolutely underestimate is networking, networking, networking. If you just go and sit behind your computer at home and you don't talk to anybody else, you can read as much as you want to online. And that's fine, but that's not going to make it.

I think people need to go and attend conferences like this. They need to go out, and they need to go and find the right communities, but also push themselves out of their comfort zones to not just go where the like-minded identical people are. They need to go and find out, so who else is out there?

So if you are somebody who is very, very good at data science, you somehow need to find somebody who's good at commercial. Try to find those people. Try to find out who else is there. I think people need to kind of go and understand that people buy from people. People invest in people. People do business with people. And therefore, we need to go and build-- the human touch is important. We need to go and build closer, still wider networks, of different people who can go and engage, help each other, because then we can build, all together, success stories.

So I think it's not so much about what do I read and what I do? It is about who do I talk to, and how open am I to just hearing different views? Because that might just trigger how to work together with and how to go and build long-lasting partnerships across boundaries, across countries, across disciplines. Because I think that's what's really needed. That's what makes the difference.

MELANIE: Any other last bits of advice that you wanted to share with our audience?

ANDREA: Just do it.


I mean, it really is--

MELANIE: Brought to you by--


ANDREA: No, no.


ANDREA: No, no, I think the--

MELANIE: Which not a bad one to be mentioning right now, frankly. Anyways--

ANDREA: No, no. But I think that the issue is, there's always a reason not to do it. If you're going to go and want to overthink things, there's always a reason not to start your own business. There's always a reason to go and play it safe. But you know what? This is the time. This is the age. There's so many opportunities at the moment out there. I think people should just-- particularly while they're young, just do it.

MELANIE: Well, Nyalleng, thank you for helping me. And Andrea, thank you. This was really wonderful. Thank you.

YASIN: This was really good, yeah.

MELANIE: Thank you.

ANDREA: Pleasure. Thank you for having me. Thank you.


MELANIE: Thank you, again, Yasin, Vukosi, and Andrea, for all coming onto the podcast to talk to us today. We really appreciate it.

Now, it's time for our question of the week. And the question of the week, main takeaways from the conference. And I'll go ahead and I'll kick this off from the standpoint of, what has been really impressive to me is to see how much-- what I see here, in terms of talking to the people that I've talked to-- is common across the board-- you know, the fears of I don't know enough, I'm scared that I-- the impostor syndrome, which everyone has. But then still, the people who are willing to take the risk to go into this space has been really impressive.

And what's even more impressive to me about this conference is how welcoming and inclusive people are. You don't always see that everywhere. And this space can definitely lead to a very, like, you don't have enough X, Y, and Z. Certain people will have that attitude. But thankfully, everyone I've come across here is just very eager to open doors, just excited to learn. And that's really amazing to me.

I'm so glad I was able to come and see that and experience that. And I'm grateful for you guys making it so easy for me to be here. So thank you.

WILLIE: I think for me, my key takeaway really is a question, which is, where do we go next? And I think, as an organization and as a community, we really have to think very hard about, where are we taking this? I mean, Africa has such a huge appetite for machine learning, for AI. There is a lot of talent here. Africa, of course, has all these challenges, which are very unique, globally. So I think we will continue to discuss what is the best way that we can really strengthen Africa and really make the world a better place.

NYALLENG: Yeah, my takeaway from this probably is just a lot of excitement, because what differentiated last year's Indaba from this year's Indaba is the leap that the last year Indaba had on the audience. Because this time around, what has happened is that people have gone back and people have gone to transform their lives. And my most inspired moment was during the Maathai Awards. Obviously, he's one of the people that is on the podcast, and I think he is inspiring.

So my takeaway from here this week is, AI in Africa is happening. And AI in Africa can have impact. And the impact, we already see the results. So there is no real need for us to doubt. Believe in the bigger things that are proposed as challenges, because for as much as they are challenges, there are so many examples now here that we have seen at the Indaba of people that have ignored those challenges and went forward and built AI and AI tools to improve African lives.

MELANIE: Yeah. Willie, let me ask you, anywhere you're going to be next, other than sleeping soon?


WILLIE: Yes, I think for the immediate future, I'm just going to go back to my own lab here at Stellenbosch. A lot of my students attended the Indaba, and I'm sure they were all really excited and inspired by everything that I've experienced. And I'm looking forward to just chatting to them and seeing where can we take our own research next.

Then of course, we also have another machine learning summer school in Stellenbosch happening in January. And yeah, we have a few exciting new programs also in development at Stellenbosch in machine learning and artificial intelligence.

MELANIE: Thank you. Nyalleng, what about you?

NYALLENG: Yes, I guess the first thing that is probably going to happen is we will probably get to Ghana. So there will be that move. And then for the actual machine learning community, I will be in Montreal in December with the Women in Machine Learning. We will be at NIPS. I am one of the core organizers of the Women in Machine Learning there, so we're running the workshop.

Yes. And I think besides that, there's going to be a month without Indaba, and so there will be a month of sleep.

MELANIE: That sounds like a good month. Well, all right. So in terms of the usual with Mark and myself, we will both be at Strange Loop next week. And so you should come and say hi to us, if you're going to be there. Outside of that, we'll share more about where we're going to be next week.

Thank you both, again. Nyalleng, Willie, I really appreciate you both joining me on the podcast and helping me put this together. Thank you.

NYALLENG: Thank you.

WILLIE: Thank you to you, Melanie.



Melanie Warrick

Continue the conversation

Leave us a comment on Google+ or Reddit