VISAGE

Every day 830 women die from preventable causes related to pregnancy and childbirth, half of them occur in Sub-Saharan Africa. VISAGE’s project aims to act on this situation by improving obstetric emergency management. Indeed, due to the lack of medical staff and diagnosis errors African women tend to refuse any tracking of their pregnancy. Proposing a phone-accessible solution, VISAGE allows women to describe, in their local language and following a pre-recorded survey, their situation. Fitting with our database, the description will determine the exit proposed by the machine that can be either an evacuation or series of advice dedicated to the patient. On a concrete level, our solution can provide a much better access to information and e-healthcare towards women who desperately need it. Indeed we can treat, on average, 10 000 calls a day and help to spot emergencies and avoid traffic jams in the medical structures. 

Maternal mortality is both a global and an African concern. Indeed, if 293 760 women died of their pregnancy in the world in 2017, 68% of them came from Sub-Saharan Africa according to the World Bank. Due to obstetric complications, venerial disease and malaria, African women face a great risk during their pregnancy. In Benin where we have started our activity it is 397 women for 100 000 inhabitants that die each year, 27% of them due to a mismanagement of obstetric setbacks. Local diseases but also a lack of tracking and of sensitization are therefore killing 548 african women on a daily basis. In l’Ouémé’s department in Benin we only find, according to the national government, four gynecologists for a female population of more than 565 000 women.  Without a central number of assistance beninese, and on a more general basis, African women can go the term of their pregnancy having not seen nor heard from a doctor or a nurse while the World Health Organization standards recommend three medical consultations to ensure a normal tracking. Therefore our solution aims to ensure the largest caring possible to reduce the number of death caused by a mismanagement of obstetric emergencies. 

By calling our robotic phone line, pregnant women can quickly and effectively describe their symptoms in their local language. Artificial intelligence is used through the process known as "speech-to-speech". It allows patients to both speak in their local language and to be understood by the software. After describing her symptoms, the patient responds to a survey, to refine the results of the algorithm. Indeed, the patient will have to enter her weight, time of pregnancy etc. As soon as all the necessary informations have been recorded, the algorithm, being the core of the solution, will match the symptoms and characteristics of the patient with standardized profiles already established by doctors with whom we work in close collaboration. Finally, our telephone line will tell the patient if her symptoms are benign or if she needs to be cared for by a medical professional. In that case, our patient’s medical file will be sent to the partner health structure through an encrypted SMS in order for  the medical teams to act more efficiently. 

Our digital solution allows pregnant women to benefit from an efficient monitoring during their pregnancy. Our main focus is on treatment of emergencies as latest reports on maternal health in West Africa show that most of the incidents occured in case of emergencies and are due to lack of coordination between services and a too-late identification of the threat. 

This solution impacts women pregnant especially in rural areas where health facilities are few and where health care workers often lack means and training.

In order to understand our target audience, we decided during the brainstorming phase to circulate in Benin in the department of Ouémé and Borgou (the test departments) surveys intended for pregnant women who could describe their experiences and difficulties they encountered during their pregnancy. We received over 150 answers.

In addition to these difficulties of access, they suffer other social and economic difficulties. Indeed, our digital solution is mainly aimed at women who are illiterate. This is why we are developing our solution around speech-to-speech technology.

It should also be noted that poverty reinforces these situations of inequality and therefore the risks of medical complications. This is  why our solution is free.

Maternal mortality in developing countries has been the cornerstone of public policies since the 1990s. However, even though progresS has been made we believe that our solution can make a significant difference especially when it comes to increasing accessibility to maternal healthcare. Indeed, by using a local language translated phone accessible solution, we participate in expanding the scope of women monitored during their pregnancy. Furthermore, if conducted in collaboration with local health structures our solution can reduce the time of  consultations and therefore help medical staff to focus on the most urgent medical issues.

Since the 2010s, the e-health market, and especially the e-maternal health market, has developed considerably. Indeed, many startups like Gifted Mum or Karangué have created innovative solutions using SMS’ in order to sensitize pregnant women to the behaviour to adopt during their pregnancy. However, regarding the management of health emergencies, there is no direct competition in Benin and in the countries in which we wish to implement our solution in a few years’ time. Indeed, the combination of artificial intelligence and a telephone line to better manage obstetric emergencies is not yet offered on the e-health market.

Our solution aims to use “speech-to-speech” technology in order to make accessible speech recognition tools in african languages to local populations. So far, the only way to use this type of communication in an African context has been through a text-based content, mostly in European languages.Therefore, we have been working on a speech-to-text solution enabling translation from african languages such as Fongbe or Yoruba toward French or English. We had the idea then to develop a solution speech to speech where what is analyse is not a text format but directly a succession of sounds. Thanks to that we reach an accuracy in understanding of Fongbe close to 99 percents. The possibilities offer by this solution are large especially because it enables our solution to be usable on a simple phone line and by people suffering illiteracy. 

Our solution is a combination of AI, big data and blockchain. We developed a speech-to-speech tool enabling an understanding of oral-based african languages. This API is connected with a second API that is a symptom checker that we built thanks to data collected in beninese and french hospitals and that is specially targeting maternal health issues.

For instance, when a call is made in Fon, a connextion request is automatically sent to the IVR server where the patient's file will be stored. Once connected to the IVR, three AI modules allow speech to speech. In addition, our software, operating from French , especially at the symptom checker level, requires a machine translator to be able to switch from French to the local language, then from the local language written to the spoken dialect.

The diagnosis is sent to the health structure where the patient will be admitted through an encrypted SMS. All over the chain the patient is anonymized and all actions taken are registered thanks to blockchain technology.

https://www.academia.edu/506455/A_robust_and_efficient_three-layered_dialogue_component_for_a_speech-to-speech_translation_system

https://www.academia.edu/33957601/Specific_features_of_the_Galician_language_and_implications_for_speech_technology_development

Our solution is currently helping 100 women to have a monitored, safer pregnancy. During this pilot phase our beneficiaries have been tracked for 4 months and then asked to evaluate the concept and especially the efficiency of the « speech-to-speech » dimension of the tool. In that regard, 4/5 of them acknowledged that the possibility to express themselves in their local language made the solution « more accessible » to the population. In addition 80% of them asserted that the exchange with our device would be their only interaction with a medical device/staff. As for now, 26 women have gone to the term of their pregnancy without detectable complications and we have suffered no hardships on a technic nor a medical level. Therefore, with very few fees as we have developed mainly the entire software, we are able to track a large proportion of women, with less than 1% of technical errors. Our trainee on the ground was mainly used to sensitize the women to operate the tool as well as for signing partnerships with clinics and maternities. With some skills and a subvention to cover our small expenditures we were able to develop a instrument easy to use and adapted to reality of the ground.  If we still need to scale-up our tests for more safety, we believe that we can provide an efficient tool to monitor obstetric emergencies. To compare our outcome with potential competitors the firm Gifted Mom helped, using no local language and throught a smartphone, 20 000 women from 2013 to 2016. At the VISAGE we believe that, with support from institutions and with a good marketing campaign, we can reach the same proportion within the next year. Indeed, if we might need some improvement on the medical part of the algorithm as well as to spend time training our field team, the fact remains that the technical part is fully operationnal and accessible and waiting for 10 000 calls (our maximum daily capability) to be realized towards our Benin’s phone line.

Currently our solution serves 100 people as we are implementing a pilot phase. In one year we aim to serve 250000 people as it is equivalent to 10% of all pregnant women each year in the countries we are targeting. According to our five year projection, we aim to serve 20 000 000 potential patients as we would like to extend our services to all kinds of emergencies, including traffic accidents that are considered as a plague in West Africa.

Our goal within the next year is to adapt our technology to languages spoken in other regions of Benin as well as in key West African countries. We are speaking about 30 languages that our solution must fully understand. In each country the same process of mapping the healthcare structure shall be repeated, as establishing several partnerships with institutions.   

Our main goal is to enable a preventive measures, and pushing people to think in terms of prevention rather than living a crisis and dealing with healthcare afterwards. 

To do so the first point for us is to develop a prediction of preventive models. Understanding which message we should convey, which tonality and to whom.

We want to change the philosophy of prevention by helping the government deliver personal messages thanks to data analysis. 

This  huge amount of data that will enable us to better understand behaviour and risks for West African women. Based on that we want to create in the next three year a tailored, fully digital microinsurance (partnership with street sellers rather than outlets). Our knowledge will help us to create it at the best price and then make it available for this amount of women. 

Our main barriers are with the collaboration with healthcare structure. Healthcare in West Africa is not free and doctors often want to maximize the profit they can make. Rather than helping us, they see our solution as a competitor that will make them lose money. Even if we convince them to become a partner, adding a new hospital to our network is time consuming. 

On a technical/legal point, West African countries’ law systems are evolving concerning rights on data. Medical data like  the one VISAGE is using can still be problematic leading us to potentially change our approach in certain countries. 

On a financial point of view, the licence for having the right to create a microinsurance scheme is expensive and takes time. Investors we spoke with say that because of the unpredictability of the process they could dismiss the investment despite the important return on said investment. We will need to complete this process in each country.  

We are currently negotiating an agreement with the Ministry of Health in order to partner with them and access their network of public hospitals. We make doctors part of the solution not through establishing a diagnosis for pregnant women, but by the information they can give us. 

Since August 2019, our team has gradually grown. Today, six people work daily on this project in parallel with their professional or student life. 

In order to ensure a better development of our solution, we have hired 5 people in Benin responsible for operations, especially to expand of our network of health structures. We also work with our key  partners who allow a dedicated team to help us reach our goals through giving us access to files, data, etc.

Finally, we have also decided to set up a strategic council composed of high level practicians and public policy experts. 

Our team is mainly composed of Sciences Po Paris students from the Europe-Africa programme, determined to put their skills at the service of this innovative project.

First of all, our project was created by the ambitions of Augustin and Baptiste Gaultier who wished to give back to Benin, as they were born there. Having already co-founded the AfricaTech UP Tour program responsible for putting digital innovation at the service of West African populations and having a personal history with maternal health issues, they decided to gather a team around them with skills diverse. They are indeed specialized in new technologies and in computer science.

American-Beninese, Imane Karimou is in charge of promoting our initiative on social networks and the web. Imane has also led other projects in favour of human development in the Republic of Benin.

Parliamentary assistant of a deputy of the foreign affairs committee  in France, Pierre Ellie is today in charge of partnerships with the private and public actors necessary for developing our device.

In addition to having skills in many fields such as tech, business and law, VISAGE has surrounded itself with professionals. A expert of telemedicine and the realities linked to African public health, Edouard Alfred Johnson, a Franco-Beninese gynecologist with 40 years of experience in France and on the African continent, is today responsible for our strategic advice. We meet regularly with the strategic council to discuss key points. 

VISAGE has set up partnerships allowing us to develop our project thanks to the partners’ expertise and advice, allowing us to technically improve our digital tool.

SciencesPo Paris: Several members of our team are still students at Sciences Po, within its program devoted to African issues.

The Clinton Foundation: Engaged in the development of innovative digital solutions in the health field, the foundation has supported us from the start by allowing us to develop through their specialists. This foundation also allows us to concretely develop our project by offering to help us in the search for grants through the Clinton Global Initiative program

UNFPA (the United Nations Population Fund): supporting us through regular meetings, UNFPA has so far shown great interest in the data that our solution could generate. We are currently negotiating a unique opportunity to use the official line of the structure to function at a national level. 

On a medical level, we called on many health professionals. Thus, as we want to first develop our tool within the issue of obstetric emergency management, we have, with the help of our Scientific Advisor, the Beninese gynecologist AE Johnson, worked together with french and beninese specialists. This has brought us to regularly monitoring the surveys and algorithms. 

We are also in direct discussion with the Beninese Ministry of Health in order to settle a partnership giving VISAGE a central point in the new health fund they are creating. 

1/ We sell our SAAS to telecom operators on a per country license basis. In a survey conducted in Cotonou, we verified that out of 100 people; ⅘ would be willing to change their telecom provider since our service is free of charge. The service will be accessible for anyone with a SIM card from a competitor, however communication costs will be higher as it is already the case today for calls between two numbers of two different operators. We are currently in discussion with MTN, the main operator in Benin. 

2/To maximize their chances of survival in an obstetric emergency, our users will be advised to create a medical file with our service. For a monthly fee, which remains completely accessible to the majority of the population, the individual will be systematically identified thanks to a card at their disposal. This will make it possible in particular to know their blood group, medical history and relatives to be notified. This user can also make available an emergency fund which, in the event of being unable to contact relatives, could save his life by allowing doctors to start the operation. This is relevant as a significant number of practitioners do not wish to start treating their patients until they have proof of payment.

3/We will also sell the data collected by developing a tailor-made expertise that we make available to insurance players. With  this offer, VISAGE is thus positioned as an entry point on a market estimated at several billion dollars. 

Our expenses have the advantage of being very low. There are telecommunications credits and access to a SAAS for contract management and team fees as we need for each language to have with us one linguist or translator able to help us reach our objective. Once the pilot phase is completed we will be able to start operating our product on a larger territory thanks to grants aiming to improve maternal health. We will also have access to grants related to technology innovation “speech to speech” process. We will start a particular premium service in which we offer people from the local diaspora to pay fees for a certain amount of women in their home region. 

However it is important to note that we estimate at no more than 3 months our market time. Yet, even during those 3 months we will be  generating revenue thanks to our tontine system.

MIT SOLVE is an incredible human adventure and participating would be a true opportunity for our team.

First of all, it would allow us to meet other solution finders and to be able to exchange with them on their ideas and their progress. We are convinced that exchanging is learning and moving forward (such as E-Eza or AcessAyfa). In addition to these valuable discussions, the feedbacks from the experts will allow us to confirm our direction, otherwise leading us to modifying our way of working and our solution.

In addition, participating in MIT SOLVE would allow us to obtain the funds necessary to develop our solution on a larger scale. This would also allow us to make ourselves known to potential investors and in particular to telephone operators who are at the heart of our business model.

As for now, our technological solution is operating concerning obstetric emergencies and COVID-19. Therefore we need to extend our solution on the larger scale of Benin throught a communication and promotion campaign. In that regard, the MIT Solve label would allow us to be more visible and credible with our partners.

Our business model is based on an exponential growth in the number of users. For this reason, we must extend our solution in a first term across Benin and then to other countries in Sub-Saharan Africa. However, we face some constraints which can be quickly resolved if we conclude the necessary partnerships.

At first, VISAGE needs a sufficient priming background to extend its solution.

We are also looking for partners specialized in communication in order to first audit our communication and then to implement an effective communication plan.

If we want to extend our solution to other countries in Sub-Saharan Africa, we need to do market research in order to develop our solution and our financial strategy fitting with the economic and social realities of the countries in which we wish to set up. 

We also need to obtain legal expertise related to data protection.

VISAGE would like to conclude partnerships with mobile telephone operators present in the different countries of Sub-Saharan Africa in which we wish to establish our solution. For example, we are currently in discussion with MTN Benin.

We also would like to conclude partnerships with national organizations. We are currently in discussion with the Ministry of Health in Bénin in order to adapt our solution to the actual green telephone line.

In addition, we would like to conclude a partnership with a legal structure in order to be able to adapt our solution to the different legalizations.

Finally, we would like to discuss with the E-Eza and AcessAyfa teams (members of MIT Solve) who are also developing solutions around maternal e-health in Sub-Saharan Africa.

VISAGE wants to apply for the Innovation for Women Prize for several reasons. We are convinced that our technology solution, based on artificial intelligence, will positively impact the lives of pregnant women in sub-Saharan Africa. Thanks to our innovative telephone line, pregnant women will be better taken care of during their pregnancy, which will prevent certain medical complications that could have been avoided.

In sub-Saharan Africa, many women, especially when they live in rural areas and / or are illiterate, do not go directly to a doctor when they feel the need. Most of the time, they do not want to incur costs in medical consultations unless there is a real emergency. We believe that we can put an end to this logic through our solution.

Moreover, this project was directly designed with women who transmitted their experiences and their needs to us through surveys. VISAGE is therefore committed to developing a digital solution that is easy to access and use for these vulnerable populations (free and based on speech to speech technology). For instance, the choice of the telephone line combining artificial and algorithmic intelligence seemed to us the most relevant.

This award will allow us to continue to develop our solution and especially to increase the number of languages understood by our telephone line, allowing more women to have access to it. Finally, this grant will permit to partially finance our communication campaigns for women in order to continue to publicize our solution.

VISAGE wants to apply to the Health Workforce Innovation Prize for several reasons. Indeed, if our technology solution, based on artificial intelligence, is primarily intended for pregnant women in sub-Saharan Africa and allows for them to be better supported during their pregnancy, our solution also aims to better support the health structures.

We are convinced that our telephone line will help to decongest health services and to improve the patient management, in particular by reducing waiting times. When our telephone line advises a pregnant woman to go to a doctor, our solution will automatically send the patient's medical file to the nearest health facility.

This award will allow us to continue to develop our innovative solution. This will allow us to finance the development of the application for these health structures. In addition, this grant will allow us to partially finance our costs in the search for partner health structures. Indeed, the establishment of partnerships requires for example legal aid. Finally, this price will allow us to partially fund the data centers during the test and development phase which will take place in September 2020.

VISAGE wants to apply for the AI Humanity Prize for several reasons. We are convinced that our technology solution, based on artificial intelligence, will positively impact the lives of pregnant women in sub-Saharan Africa. Thanks to our innovative telephone line, pregnant women will be better taken care of during their pregnancy, which will prevent certain medical complications that could be avoided.

We are thus developing a telephone line based on artificial intelligence and in particular on speech to speech technology which allows pregnant women to describe the symptoms they encounter orally and in their local language. Through an algorithm designed with health professionals, our telephone line tells the pregnant woman if her symptoms are mild, in this case preventive messages are sent to her, or if this requires a medical consultation. In that case she is immediately referred to the nearest health facility where her file is sent.

Our solution is a combination of AI, big data and blockchain. We developed this thanks to artificial intelligence a speech to speech tool enabling an understanding of oral-based african languages. This API is connected with a second API that is a symptom checker that we built thanks to data collected in beninese and french hospitals and that is specially targeting maternal health issues. The diagnosis is send to the health structure where the patient will be admitted through a encrypted SMS. All over the chain the patient is anonymized and all actions taken are registered thanks to blockchain technology. 

This award will allow us to continue to develop our solution and especially to increase the number of languages understood by our telephone line and therefore allow more women to have access to it. This award will also allow us to continue to develop our solution. This will allow us to finance the development of the application for these health structures.

VISAGE is applying to The Bill and Melinda Foundation Award for multiple reasons. We are convinced that our technology-based solution, using artificial intelligence, will positively impact the lives of pregnant women in sub-Saharan Africa. They will be better supported during their pregnancy, which could prevent medical complications.

In sub-Saharan Africa, many women, especially when they live in rural areas and / or are illiterate, do not straight-away go to the doctor when they feel the need to. Most of the time, they do not want to sustain costs in medical consultations unless for real emergency. We believe that we can put an end to this logic through our innovative telephone line.

We are thus developing a telephone line based on speech to speech technology which enables pregnant women to describe their symptoms orally and in their local language. Through an algorithm designed with health professionals, our telephone line tells the pregnant woman if her symptoms are mild, in this case preventive messages on the actions she should take are sent to her, or otherwise this requires a medical consultation. In which case she is immediately referred to the nearest health facility where her file was sent.

Moreover, this project was directly designed with women who testified their experiences and their needs through our surveys. VISAGE is therefore committed to developing a digital solution that is easy to access and to use for these vulnerable populations. The choice of the telephone line combining artificial and algorithmic intelligence seemed to us the most relevant.

We are currently in discussions with the Ministry of Health of Benin in order to set up a lasting partnership between the Beninese public administration and the VISAGE team. Indeed, in accordance with the will of the Beninese President Patrice Talon, we have the same objectives of reducing maternal and infant mortality for Benin. The partnership project would be to take over the green telephone line for pregnant women in order to integrate our technology. There are today a dozen telephone operators for the telephone set. By adapting the telephone line to our solution, it will be able to handle more than 10,000 calls in one day.

This prize allows us to continue to develop our solution in Benin as well as in other countries in sub-Saharan Africa such as Côte d'Ivoire, Ghana or Nigeria. It gives us some credibility with our partners who may be reluctant. This also allows us to obtain the number of languages understood by our telephone line and therefore allows for even more women to have access to this service. This grant also allows us to partially finance our advertising campaigns aimed at women in order to publicize our solution. Finally, this prize will allow us to partially fund the data centers during the test and development phase which will take place in September 2020.