How AI, machine learning are being increasingly used to drive social good

Data and technology solutions are key to reaching the most vulnerable children at speed

While technologies like artificial intelligence (AI) and machine learning (ML) have become part of many aspects of our daily lives in recent times, they are increasingly being used for the social good as well. 

Take the case of the Foundation to Educate Girls Globally (FEGG), a not-for-profit organisation that aims to find solutions to enrollment and learning gaps in primary education. The foundation identifies girls who are out of school, enrols them and makes sure that they continue going to school.

FEGG, which has been in existence for more than a decade, collects data on out-of-school girls by conducting door-to-door surveys at the village level. For the last two years, it has taken the help of IDinsights, a global data analytics and research organisation, to use AI and ML for this exercise. Since then, the accuracy of identification of the villages that have the highest number of girls out of school has gone up by over 80 per cent.

“In 2015 we launched the world’s first development impact bond, where, with IDinsights, we used tech to conduct impact evaluation. After that we realised that we can use AI for identifying potential villages with a high number of girls out of school,” says Maharshi Vaishnav, FEGG’s chief operating officer. 

Operating in the states of Rajasthan, Madhya Pradesh and Uttar Pradesh, FEGG has so far obtained data on out-of-school girls in 50-55 lakh households. The algorithm created by IDinsights takes FEGG’s data, overlays it with the publicly available government data, and runs exclusion parameters against both sets of data. The system filters out those villages from the government data that have a higher propensity of out-of-school girls. 

“This has brought down the time taken for surveys and their cost by 50 per cent. The algorithm predicts the districts and villages that have a higher number of out-of-school girls, and hence we only visit those villages. The precision level is as high as 88 per cent,” says Vaishnav. 

This has also led to a rise in the number of enrollments of girls. In the year before the pandemic, FEGG was able to reach out to 200,000 girls out of school. Going ahead, the foundation wants to come up with a vulnerability index that allows them to link these girls to the various government schemes for the welfare of the girl child.

Rocket Learning is another not-for-profit that is using AI to drive its social mission. The edtech startup uses AI to make parents from low-income groups a part of their children’s learning activity. Since its inception in 2020, Rocket Learning has touched the lives of one million children in the age group of 3-8 years, and enrolled them in anganwadis and government schools.

Aiming to develop early childhood and foundational learning in the bottom percentile of the population, Rocket Learning’s unique selling proposition is to use technology to make parents participate in this journey, especially during the pandemic, when schools have been closed and parents and teachers have struggled to adjust to online classes.

Rocket Learning has designed a tech platform that creates digital communities of parents and teachers. The educational content consists of simple audio-visual activities the child needs to perform with the help of a parent. Automation content is sent on WhatsApp daily and Rocket Learning asks parents to send back the responses, thus involving them in the learning exercise. The system also sends out real-time nudges and uses data analytics on parents’ feedback. 

Vishal Sunil, co-founder and chief technology officer, Rocket Learning, believes that what really worked in their case was the asynchronous approach to the learning model. “Most online education models at the government level did not work, as they use synchronous teaching modules and online classes,” says Sunil. “The other big issue is, how do you keep the parent motivated to send the response back? In our model, we have seen 40-50 per cent parents actually engaging with the content.”   

Experts reckon that with Big Tech players like Google opening up source codes for the developer community, AI and other deep-tech tools have become accessible, in terms of both cost (free) and ease of use. Hence, these technologies are now being used for a variety of programmes for the public good.

Health care, for example, is a segment that has several use-cases of AI, especially in the ongoing battle against Covid-19. “We worked with several state governments and city municipalities to better understand contact tracing to control the spread of Covid. Our company’s proprietary platform, which used several AI models and maps, helped the government in effective contact tracing models,” says Akash Bhatia, co-founder and chief executive officer of Infinite Analytics. 

For the US market, Bhatia’s team created a well-being index, which can identify communities that need vaccination on a priority basis. “It’s an example of how AI can be used to understand people’s behaviour patterns and health parameters in any pandemic-like situation,” Bhatia adds.

The Indian government, too, has made AI and ML part of its effort to fight Covid-19, by using tele-health, tele-education, and so on. 

One of the concerns of using AI and ML in the social sector is that the data being collected may never have the consent of the user. “In our case, the algorithm does not need personalised information — it only predicts at a general population level. But yes, the moment we start tagging these girls to government schemes, we will have to create unique identifications,” says Vaishnav.

Vaishnav also points out that if the United Nations’ sustainable developement goals are to be met by 2030, then use of data and technology solutions are key to accelerating the journey towards achieving these goals and reaching the most vulnerable children at speed.

“AI is among the most exciting emerging technologies around, with great potential to leave a lasting impact on society,” said Debjani Ghosh, president, Nasscom. “In the past decade, India has slowly but surely cemented herself as an AI hub, fostering deep-tech startups, promoting academic research and enabling digital transformation across every major sector. AI is not only providing immense value to businesses but equally importantly, is addressing the social good effectively.”