AI Hero: Kriti
Kriti grew up in the state of Rajasthan, in northern India. She had a happy childhood with supportive parents who encouraged her to learn and be independent. But even with that kind of encouragement, nobody would have guessed then that Kriti was going to be recognized later in her life as one of the 30 under 30 by Forbes for advancements in AI.
Like many other AI heroes you will get to know in this book, Kriti’s attitude toward her constraints defined her. She didn’t have access to a computer or the internet beyond the 10 computers available in her school for thousands of students, but she did have access to the library, which opened up an entire world of knowledge for her. At the age of 12 she built her own computer, putting together different parts she’d acquired for a few dollars.
When she was still a teenager, she was invited to visit a government space research lab. Kriti was fascinated by space exploration and how computers can help us analyze the huge amount of data gathered about the universe to understand it better. But that wasn’t what captured her imagination the most: it was how a particular researcher was using one of the computers that changed her life. Instead of a keyboard or mouse, this researcher was using voice commands to control their computer. From that moment, she knew she wanted one of those computers (at that time the size of a room!) for herself.
It was the same independent streak and desire to learn that took her to university—a remarkable occurrence, as most of Kriti’s peers instead were destined for marriage at a young age. The median age at first marriage for women in India is 16.4 years, and in some areas female literacy is as low as 50%. Those numbers, among others, would come back to haunt Kriti years after she left India.
After graduating from the University of Rajasthan, she received a full scholarship to the University of St. Andrews in Scotland, where she obtained a master’s degree in advanced computer science. She was about to accept another full scholarship for a PhD in Oxford when she realized she couldn’t wait any longer—it was time to stop studying and start applying AI in the real world.
Kriti’s first job was at Barclays. Like many other industries, banking has experienced a huge transformation driven by technology and software. Kriti’s role was to transform the bank once again, this time to make it a data company. As you’ll learn in the next chapter, that transformation can start by focusing on existing business applications and how to make them better with AI. For example, Kriti’s team modernized the existing rules-based applications for financial transactions and applied modern AI-driven techniques instead, dramatically improving the customer experience.
In her next job, Kriti concentrated on conversational AI. Technology that had once only been available at research institutions like the one she’d visited as a schoolgirl was now available to any computer or smartphone user. She was fascinated by how natural conversational interfaces can make technology more accessible and easier to use; that’s not only cool, but also has profound implications for the democratization of technology. Instead of humans learning how to communicate with computers, computers can learn how to communicate with us.
Kriti joined Sage, one of the largest business applications companies in the world, to pursue that goal. Sage’s main customers are not big corporations with ranks of accountants and tax experts. Most of their customers are small businesses that struggle to keep up with the administrative work. They spend an average of 120 days a year on administrative tasks, which takes a huge toll on their productivity and stops them from being proactive and forward-looking, limiting growth in what is the backbone of many economies in the world.
As Sage’s VP of AI, Kriti began working to change that. In 2016 she helped create Pegg, one of the first business-oriented conversational AI agents in the market. Pegg was intended to help small companies manage their finances through a chat-based interface available in popular messaging apps like Slack, Skype, and Facebook Messenger. It makes administrative tasks such as uploading a receipt, logging a payment, or checking a balance as simple as texting a friend. On the backend, Pegg uses AI to help business owners without accounting experience to be proactive with their finances. It can make recommendations, send notifications for accounts about to go past due, and provide future projections.
Pegg is a great example of how AI can democratize technology to underserved audiences—in this case, small businesses. But this was just the beginning for Kriti. Her next project, leveraging the same technologies she was using to transform corporations, would be motivated by her memories of her childhood in India.
Growing up in India as a woman was very different from Kriti’s new life in London. India is considered to be the world’s most dangerous country for sexual violence against women. A crime against a woman is committed in India every 3 minutes, and every 20 minutes a woman is raped. The problem is deeply rooted in society. A staggering 65% of Indian men believe women should tolerate violence and that they sometimes deserve to be beaten. Some 24% of men also admit they have committed sexual violence at some point during their lives.
In such an environment, it’s not difficult to imagine the constant fear that many parents of girls in India suffer. Kriti still remembers how her parents wouldn’t let her go outside by herself. She was lucky, though: some Indian parents do not even want a daughter in the first place. Selective abortions and even female infanticide are the most horrible consequences of gender inequality: in Rajasthan, the state where Kriti grew up, there are 888 girls for every 1,000 boys under the age of 6.
It was during a dinner with a group of women in South Africa when that reality came back to Kriti’s mind. It was a regular dinner, but the casual nature of it added a halo of safety that led one of the women to share about suffering domestic violence at home. Suddenly, more of the women around the table began sharing their experiences. Most of them had suffered at least one episode of sexual, physical, or psychological abuse.
Kriti was shocked by these stories, but unfortunately the chances of finding several women with past experiences of abuse in South Africa are quite high. One in three women face physical, sexual, or psychological violence at home. In most cases, the violence stays within those walls; only 5% to 10% of such incidents are ever reported, and when they are, it’s usually too late. The average number of episodes of abuse a woman suffers before asking the authorities for help is 35. The first 34 times she keeps silent, often because of the social stigma or just shame of admitting having been abused by her own partner.
Kriti was not an expert in such a difficult problem, nor did she think she could fix it completely. But she knew technology could help. Working with local institutions, she created rAInbow, a digital companion for women at risk of abuse. Research and interviews with abuse victims have shown that rAInbow is able to help women in toxic relationships realize they are being abused and take action. Not having a real person on the other side actually makes the system more approachable: conversations with rAInbow are nonjudgmental and unbiased as well as emotionally aware.
Just as Pegg has an AI backend that can provide proactive insights about finance, rAInbow can identify early signs of abuse and be proactive. It has been trained to recognize behavior patterns leading to abuse, inform victims of their rights, and encourage them to seek assistance.
rAInbow was launched in 2018 in partnership with Nelson Mandela’s stepdaughter Josina Machel, who had also lived through a horrific domestic violence experience that left her blind in one eye. Since then, hundreds of thousands of women have used the service, and it has made a huge difference in many of their lives. rAInbow was recently recognized as one of the top digital innovations of the year by UNESCO.
After working on rAInbow, Kriti kept identifying social issues that could be addressed with technology. She didn’t have to search far, as things in her native hometown had not changed much since she’d left. Gender disparity is deeply rooted in Indian society, and from early childhood women are raised in a culture in which information is withheld by adults. Sex education is a taboo, and girls often have to face their sexual maturity without any help: 9 out of 10 girls in India don’t know what their period is, and many think they are dying when it first happens to them. They don’t have any information on birth control and they are not aware of the risks of sexually transmitted diseases.
With the help of the Population Foundation of India, Kriti launched another digital assistant called Dr. Sneha in early 2019. Like rAInbow, Dr. Sneha is a friendly digital companion with whom users can have unbiased, nonjudgmental conversations; girls can feel safe asking Dr. Sneha the questions they cannot ask the adults around them. The service can be used even on very low-cost phones, and it is distributed through community centers in rural areas where girls don’t have access to mobile devices. The interface is also adapted for the low literacy rates still present in some of those areas, using visual cues and video storytelling instead of plain text. In its first month, it had already provided one million sexual-health consultations to young people in India.
As I was writing this book, Kriti had just left her leadership position at Sage to fully dedicate her career to the use of AI for social issues at AI for Good, an organization she founded with the mission of creating technologies for solving some of the toughest challenges facing humanity. In speaking to her, you can feel Kriti’s excitement for the journey ahead. But that excitement has little to do with creating new breakthroughs in technology, or advancements in AI. Actually, some of Kriti’s biggest accomplishments didn’t use the most complex or cutting-edge AI techniques; she just used core AI technologies to create better software (which you will learn to do in the next chapter). Her excitement comes from applying this new breed of software to real problems. It comes from combining the scale of software with the human-centric possibilities of AI to create products that can help millions of abuse victims to open up and look for help, or girls to learn about sexual and reproductive health.
The world of the future will need AI leaders like Kriti. More than ever, technology will need to be more human, and the people behind it will need to recognize that.