- 6 hours ago
Most health tech is built for systems that already work — stable infrastructure, funded hospitals, digitised records. Shikoh Gitau, Founder and CEO of Qhala, has spent her career doing the inverse: building health solutions from the ground up when the infrastructure, the regulation, and the user behaviour are all still taking shape at the same time. In this conversation, Shikoh shares what patient-centred design looks like in rural areas, how you influence policy as a founder, and what Africa's health tech moment tells us about the future of innovation beyond Silicon Valley.
Category
🤖
TechTranscript
00:00welcome to them but hello how's everyone doing we're feeling good right yeah feeling some good
00:24energy i think the husband is there hello yeah we'll see okay let me just put this water bottle
00:30away so this is an amazing stage i've got to say it's my first turn here it's kind of cool
00:37right
00:37so as you can see i'm standing here with the founder and ceo of carla she could
00:44sorry gtao she gtao so we know that of course everyone today everyone here at vivatech they're
00:51all talking about ai they're talking about efficiency but in this conversation we're
00:57talking about ai done differently ai done by african companies by african people for their
01:07people the people that understand the roots of where they come from hello sam now africa's
01:15healthcare is decentralized it lacks hdps we know this health infrastructure and ai is seen
01:21there as a force multiplier something for the good now carla wants to ignite africa's digital
01:28transformation especially to act as a game changer in healthcare so it's an absolute pleasure to
01:33welcome you shiko to our black stage so talk us through while i try to sit on this um i
01:39will stand
01:39for a second because a little bit unladylight with the heels um talk us a little bit about where carla
01:44comes from now we were talking backstage and i was saying actually carla in urdu so the pakistani
01:50language it means black so we're on the black stage so welcome um but you were also saying
01:56that in zulu it means to spark yes so take us through what it's all about so um so thank
02:03you
02:03very much i only know two words in french it's bonjour and how do you say straight address oh uh
02:11street straight oh straight um tudua tudua i used to live in court d'ivoire i only knew how to
02:17give
02:17directions to the taxi to get me home that's the only french i know but thank you so much for
02:22having
02:22me this is such fun and yes uh color means to spark or to ignite in in zulu and our
02:30goal is to catalyze
02:31africa's digital future and why why health for us um so i always like giving numbers because it's
02:37important to paint context yeah in every given african country uh the ratio of doctor to patient
02:44is quite depressing and i'll pick kenya and kenya is quite developed compared to many other african
02:51countries there's one doctor to 17 patients 17 000 patients one person to one doctor yeah
03:01you see where we are going and let me even make it more blake right there are only three thousand
03:07one two yes three thousand pathologists on the whole of the african continent
03:14is that sinking in yeah so when you think about ai it's a life and death situation when it comes
03:21to
03:21health for you to get a cancer diagnosis given that ratio of a pathologist it will take between 50 i
03:29mean
03:29between 50 and 70 days by that time if you had like an aggressive cancer you're dead
03:34yeah uh most of the health um authorities require seven days of diagnosis so so for us it is important
03:42and it's actually important for for us to be able to take advantage of this moment and see how it
03:47works
03:48for the continent and if you solve for health you solve for a majority of the problems that many people
03:52have because they're able to invest that money in other areas of their economic um development
03:58now your background is a computer scientist so what was it about you know did you basically see
04:06the broken health care system and say that i can really make a change here was that the sort of
04:12ambition behind it all yeah so i keep joking that i went on as a kid when you know every
04:16kid even
04:17you you are asked which way you want to be when you grow up and in africa you are you
04:22are either an
04:22engineer a medical doctor a lawyer or a teacher you didn't have like a variety of careers to choose
04:28from when i was growing up so i used to tell people i wanted to be a medical doctor i
04:32never ended up being
04:33one so it's a passion i had but during covid we came by uh health during covid we were a
04:41group of data
04:41scientists and engineers and we were told to solve for health using data because covid the way it was
04:49spreading it was not just a health problem it was an economic problem it was a socio-economic
04:56problem it had many facets to it and it became this kankwark for the for the for for for for
05:01africa
05:02basically because it's up a number of countries it's in the process of doing that and seeing just
05:06the impact of covid on livelihoods that drove us to want to solve for health people who were like
05:13middle income and middle middle income went to poverty because one of their life loved ones needed
05:19oxygen and it just became a matter of like how do we start solving for health in a more proactive
05:27manner using technology now you're mentioning a stat there thousands of people for one doctor so talk
05:34us through how carla actually works because um i would say it's bold it's inclusive but can
05:41the way in which you design the technology actually be scaled up to tell the audience about it so i
05:47mean i've painted such a bleak picture but there's great advantages on the african continent every
05:53household has a mobile phone yeah and a smartphone to that to that extent so every time you see news
05:59don't believe all of it yeah most households have mobile phones that are smart that are connected to
06:05the internet we have we are the highest per capita users of mobile money digital currency is mainstay
06:14in kenya for example and most of africa so we don't need an actual physical bank to do anything so
06:20we
06:20have all these things that are working so how do you use the assets that we already have to provide
06:24access to these critical services like medical health so for example i'll give you an example they
06:30there are people who are called buda buda riders buda buda riders are found in every city in africa
06:35and in india i found out because this there are these motorcycle riders that that carry people
06:40when you come to nairobi or you're in kampala or kigali you'll find them right especially kampala
06:47they're everywhere and these are your always your first response when in any medical emergency
06:53whether in the city or in the village they're the people who are going to be called to actually
06:57provide that so who should be empowered to provide medical care it's these people we don't have
07:03enough doctors right we don't have enough committed health workers how do you empower somebody who's
07:08never gone to school because most of them have not gone to school is you provide them with a tool
07:14to be able to that will engage with them at the level that they know in their local language and
07:19that's what ai provides is i click i'm chatting with a whatsapp bot in my language and i'm not
07:25chatting i am having a conversation a voice fast conversation my language fast conversation with
07:31these people so we've just rolled a pilot of this in kenya with buda buda riders and running it out
07:37what happened next is they said okay we have this solution but these people when they go to hospital
07:41they will die because they don't have blood i have a mother who i've taken to hospital giving birth
07:45they died because they didn't have enough blood they started organizing themselves to have blood
07:50dries are you seeing that's how ai can push people to action because they get empowered and they start
07:56solving problems for themselves and that's what i love about ai it's it's giving people the the
08:02dignity to know that they can contribute to their society and that's that's what we're trying to do
08:07yeah it's amazing um now you brought up childbirth i mean obviously we've all read stats we know what
08:14happens because especially in these rural communities these far-flung communities um they can't women
08:20can't access or their families can't access um medical health and then these women die in childbirth so
08:26how does your technology how does kala help provide the medical aid and attention that they need so that
08:33these women and their children survive awesome so as i mentioned our we build tools that are trained
08:41on local medical guidelines it is not something that we pulled from from america or from europe
08:47it's it's trained as a kenyan doctor yeah or kenyan nurse and so we ensure that this model is really
08:56good enough to be able to walk somebody through the process of childbirth or if they see a woman they
09:02have swollen feet they're having a headache they say okay that's preeclampsia it needs actual urgent
09:07health care and being able to say give right information when it's needed but also connect
09:13directly to our human being a medical doctor a community health worker somebody who is able to
09:19quickly give guidance when it's needed and so and saving lives 40 percent of women who die
09:25from childbirth it's it's situations that could be avoided if care is given at the right time and
09:32that's what we do is we train models small language models not and they can run on edge yes because
09:39everybody's going to ask me about compute they can run on edge they are small models they are able to
09:45be able to be pushed yeah to women to know to anybody actually but we're primarily looking at people
09:51who are fast fast uh fast fast responders and then the first responders are able to provide that care
09:57before you have to rush a woman to hospital you rush a woman to hospital when they are stable
10:01you don't want a woman dying on the way to hospital you provide care when they need it
10:06yeah so you've written in the natural environment you could say into the code but what about the end
10:12user how are they experiencing it because do they need to have the latest iphone or the latest
10:17something no right so how does it work for them you need whatsapp that is what you just need whatsapp
10:21and we all know whatsapp runs on any phone probably it can connect onto the internet right
10:25and it's whatsapp bots that are able to have a conversation with you anyway ensuring that it
10:31can in the very basic version of it speak the widely most used language in that country and in kenya
10:37for example it can speak swahili and it's normal swahili not latin swahili it's normal swahili we are
10:42having a conversation with this board it's able to tell me what are you seeing in swahili okay when you
10:49see this this is what this is the first step you need to take have you taken that step have
10:53you
10:53taken the next step do you want to show me to show to see this in a video form or
10:59a photo form it's
11:00multi-model so that it's providing the best support to somebody who is like totally starting off this
11:06process or even somebody who is quite advanced we have these models being used by community health
11:12workers who are well trained yeah to provide care um in the in the villages yeah now i've had a
11:18few
11:18conversations across other stages where we've discussed you know the way in which people are
11:22using ai and so if people are using it for everyday purposes so in your case healthcare should we
11:29humanize the ai i don't want to ask you that question but what i do want to ask you is
11:34whether
11:34or not so you've said that you know you've built in baked into the code um all of the different
11:39kind
11:39of languages you've learnt you know it understands how to communicate properly and easily with people
11:45um on whatsapp but is is is there some way where you feel as if the ai could do better
11:51is it talking
11:52to them in a way that that so it humanizes the conversation and and they feel that they can trust
11:58and that they are cared for through this so one of the key things so i mean i've talked about
12:04the
12:04application of some of the things that you're doing as i mentioned one of the key things that we are
12:07really hard working on is the context yeah so right now yes you're using the health guidelines from the
12:13from the government but there's traditional um what is it called tradition uh information that
12:20is lost yeah because we are using this model but there are women in in the village who've been giving
12:25birth for hundreds of years who know how to do some of these things there is medicinal plants that
12:30are available in in that village that is able to take care of the stuff that we are we are
12:35trying to
12:36take care of right now is how do we bring some of that knowledge into the models and that's why
12:41we say
12:41africa is only contributing two percent of the data that is training most of the frontier models
12:46is how do we make sure that the continent is well represented and we're not losing our knowledge on
12:51our our contextual knowledge but also our indigenous knowledge and our and our cultural knowledge that
12:58will help us as a continent move forward yeah so you you managed to gain some of the indigenous
13:04knowledge through your um through through your application but have you ensured that there is
13:11a sort of backstop so that other people can't access that data and information because i mean we we were
13:17speaking about it earlier i mean we won't go into it too much because it takes away from the positivity
13:22and the work that you're doing but how do you ensure that your data is safe yeah and and and
13:26that's a big
13:27question right now because as i said the one advantage that africa has is that 98 percent of its data
13:32is not available to the world yeah it's it's a disadvantage and an advantage at the same time
13:38because we we get like we have some cards on our chest right that we can trade in but how
13:44do you stop
13:44it from being taken advantage of how do you stop it from being uh misused and how do you stop
13:52africa from
13:53staying back in the innovation space yeah so many people are putting many things in place like this
13:58licensing that is coming together that is just being able to license the data that is being
14:03collected by indigenous people such that if you're going to access it can you please make the the
14:09access to the code not a hundred dollars a month please yeah make it accessible faster and better to
14:15the people who are providing that data so those are the negotiations and conversations that are
14:18happening right now is that if you're going to access our data to to create a model that is a
14:23multi-trillion dollar at the moment is can we have access to that data at a cheaper process
14:29or can i license and sell you the data it is a conviction that is happening rapidly and because
14:34we have lots of data that is not ready yet we have something actually to offer to the world
14:39now considering that you're gaining some very valuable data you know when it comes to
14:43healthcare and then we we speak about you know the future of healthcare we're on the kind of
14:47oh sorry i should almost drop there um you know we're on the stage that's all about the
14:51innovation healthcare so at the moment what your tech does is treat people i mean obviously you're
14:56you're treating people in communities that have not had access to healthcare in a while but can
15:01you see a future where the data data that you have the access that you have to the kind of
15:06communities that you have can help one day lead to preventative care as well so that's actually
15:12it's not a future thing it's happening now it's happening now okay amazing so so we we keep
15:18thinking about like biotech and stuff and biotech becomes an american biohackers i mean we've been
15:24reading about all these people peptides if you're not using peptides and you're here
15:27yeah anyway i can i can i know people are smiling because they are they're deep in the in in
15:32in it
15:33but when you when you when you read about these things and the science behind them it has taken
15:38very very little consideration of the african and the african human body yeah right now i think i can
15:44talk about it so we we've been working on this project for the last year and a half and it's
15:48about
15:49women reproductive for for african women do you know how many uh reproductive health issues that an
15:56african woman has how do you measure them how do you compel policymakers to make a decision on data
16:03so we are building the world's first and i can say that gladly the world's first uh utero index and
16:10it's basically measuring what are the different issues that a woman in africa actually faces from
16:15the the moment their body switched on to when they are post-menopausal and and measuring them and
16:22showing doctors this is what you need to be taking care of because you can say i have pico pcos
16:27or
16:28pmos nowadays but a pmos of a woman in africa shows very differently yeah than a woman here so if
16:35you're
16:35treating something here and you're trying to use the same guidelines and a protocol in a for a woman
16:41in africa it would not work so some that's some of the work that you're working on actively and we
16:45started with women because that my my friend fatru and i are both women and we face like really high
16:50reproductive health issues so we're saying we're solving for ourselves but also we are solving for
16:54the millions of african women who did not have the kind of access we had to go to solve these
16:59issues
16:59for themselves yeah now we've got a couple of minutes left so i want to ask you the question that
17:04everyone always ends up getting into which is the kind of regulatory landscape so let's take that
17:10let's let's talk about regulation in a way that you know is is a little bit more positive i mean
17:15where can regulation be the driver for innovation in health care um and for others to join you in
17:24this quest as well so when i think about regulation i think it in four boxes uh box number one
17:31is like
17:31market fast that is where where where the us is it's driving uh towards like let's move fast break
17:38everything and let's no no no no no one will help back and then there's a europe one which is
17:45like
17:45it's you're protecting your market let's slow down on this innovation thing let's think it through
17:51and there's china one which is like in the middle of both of them and there's indian us where we're
17:56saying we are holding we are observing as we are progressing yeah and that's where we should be
18:03regulation should be saying if we are observing we acknowledge that this is a very strong and could
18:08be harmful technology but we also know we need it as i mentioned in health we actually need it in
18:13education in agriculture we need this technology for our continent but let's put guardrails
18:19not regulate yes okay so we're obviously here at vivotech um where do you want if we're sitting
18:31here at vivotech in maybe five years times because obviously everyone wants to hear from shiko again
18:36what would success look like for you and then also success for the people that you are helping today
18:44so right now when you speak about ai in africa everybody fronts fast is like there is ai in
18:50africa yes there is what i want people here is to acknowledge that there's ai beyond your chat gpt
18:56your ability to create uh decks or vibe code nowadays the biggest one is vibe code your ability to vibe
19:03code there's ai that is being used to actually uh save human human life and maybe some of the
19:09technology we're building can be brought here in the in europe yeah or be taken somewhere else where
19:14it's we call it practical ai it's ai that is actually moving the needle in people's lives beyond
19:20productivity and that's what i want people to be able to acknowledge and hopefully we'll be talking
19:24beyond claude and and fable and and all these more models to how do we actually change lives using
19:32this powerful technology well listen chico um it's been a pleasure and i hope for the audience you've
19:38understood that there's a much better conversation to be having about the innovation that africa is
19:45leading front and center thank you so much thank you everyone round up for chico please
Comments