AI Assisted Healthcare Workers with Myna Bolo (‘Speak’ in Hindi)

Women in India still face barriers to accessing sexual and reproductive health (SRH) services due to stigma, discrimination, and lack of high-quality options. The ability of these girls and women to protect and foster their own SRH needs is hampered by a huge gap in knowledge about their health, bias among health providers, and social stigma. Over 3.4 million women in India have a desire to avoid pregnancy, of whom 94% are married. Moreover, 71% of girls report not even knowing about menstruation before their first period. As women get more educated, stay in school for longer, and aspire to go to college and work beyond, there is a dearth of programs focused on them in this critical stage where their trade-offs with other aspects of life start to increase. To compound these challenges, women living in urban slums have less space and privacy than their rural counterparts, which further hinders their ability to take care of SRH concerns. 

Most women in India have historically learned about their menstrual cycles and other sexual health queries from their peers, who themselves have often learned through rumors, WhatsApp forwards, or misinformation on the internet. This perpetuates lack of knowledge, continued stigma, and the fear of being judged by family-connection health providers. Community members perpetuate taboos and misconceptions about menstruation and SRH.  Women in urban slums have little space and privacy, and are overloaded by a variety of misinformation from competing sources; comparatively, women in rural areas have less access to health information overall and are less exposed to newer norms. 

Hence, women need an unbiased, factual source of information about their sexual and reproductive health that comes from a woman in a non-judgemental, confidential, accurate, and actionable way given their social norms and existing context. The advice also needs to be personalized for them given their specific conditions with references to local female doctors for prescriptions and medical treatments. Further, how information is communicated is just as important as what is conveyed if we want to make sure they actually understand and benefit from the advice. In urban slums where there is widespread mistrust, reliable and empathetic human sources of information are key to success, which is where our solution comes into the picture. 

Our project, centered on the Myna Bolo chatbot, is designed to bridge a critical gap in AI technology, especially in serving women from India's underserved communities. Current LLMs struggle to meet the unique needs of women, lacking in recognizing local dialects, cultural nuances, and behavioral patterns, especially for marginalized women less represented online, thereby increasing the AI technology access gap. 

Our solution is to deploy a comprehensive series of training courses  titled "Generative AI for AI-Assisted Healthcare Workers." This initiative aims to combat misconceptions around women's health by establishing an AI-powered Healthcare Women Workforce in India. Our other project, the RANI initiative, encompasses a workforce structured around two pivotal roles: Healthcare Prompt Engineers (HPE) and AI-Assisted Healthcare Specialists (AIHS). These roles, undertaken by RANI workers, come equipped with unique skill sets and tailored modules that drive our mission forward. RANI workers specialize in healthcare- related tasks, gaining knowledge about generating medical inquiries and responses, bolstered by Myna Bolo, which aids in addressing complex issues surrounding women's sexual and reproductive health. 

Healthcare Prompt Engineers (HPE) will focus on crafting effective prompts that elicit optimal queries from GPT, specifically within a healthcare context. They also receive training in medical imaging and complex medical terminology. These HPEs play a vital role in generating essential content for localized generative AI LLM models. On the other hand, AI-Assisted Healthcare Specialists (AIHS) specialize in medical literature, best practices, and dispelling common misconceptions about women's bodies. They master the art of using AI to provide personalized, accurate responses, enhancing their own empathetic and intellectual skills with medically precise AI-driven insights. For domain specific knowledge, we work closely with a network of doctors and nurses who provide expertise to validate responses and create the training materials. 

The project operates within a community-driven feedback loop, with local women (RANI workers), trained as Prompt Engineers, inputting healthcare-related queries and responses refining the LLM model for AIHS Specialists. The Prompt Engineering courses cover data pre-processing, language and content generation, and AI output evaluation. Specialists focus on dispelling social stigma around women's bodies using AI and their medical knowledge. Our innovative solution pioneers an AI-driven, women-centric approach to address misconceptions and improve healthcare knowledge.

Myna Bolo's solution is poised to revolutionize the lives of millions marginalized women and girls in urban slums across Mumbai and beyond, addressing critical gaps in sexual and reproductive health (SRH) education and providing dignified employment opportunities .Our mission extends to major cities like Ahmedabad, Bangalore, Delhi, and Lucknow, using generative AI to empower underserved women, fostering equitable growth and transformative change.

For countless women in these communities, menstruation has long been a taboo subject, shrouded in shame and misinformation. The lack of accurate SRH education perpetuates cycles of ignorance and myths that harm women's health and well-being. By providing on-demand, culturally relevant information,  Myna Bolo ensures that women have access to accurate, personalized advice regarding their SRH concerns. This empowers them to make informed decisions about their health, breaking down barriers of stigma and secrecy.

Moreover, Myna Bolo's innovative approach to employment creates pathways for economic empowerment among women who have long been sidelined by societal norms. By training women as Healthcare Prompt Engineers and AI-Assisted Healthcare Specialists, Myna Bolo not only equips them with valuable skills in technology but also offers them flexible, remote job opportunities that accommodate their familial responsibilities. This enables women to contribute financially to their households while also prioritizing their own personal and professional growth.

Crucially, by becoming trusted sources of SRH information within their communities, these women serve as catalysts for change, challenging entrenched beliefs and advocating for evidence-based practices. Through peer-to-peer education and community outreach, they help dismantle harmful misconceptions surrounding menstruation and reproductive health, empowering future generations of women to take control of their own bodies and destinies.

In essence, Myna Bolo's solution represents a multifaceted approach to health equity, addressing not only the immediate SRH needs of marginalized women but also tackling systemic issues of gender inequality and social stigma associated with it.

Myna Mahila Foundation and its team are uniquely positioned to design and deliver the Myna Bolo solution to marginalized communities in Mumbai's urban slums. Led by Suhani Jalota, who has deep roots in these communities, the foundation has fostered trust and collaboration over its eight years of operation. Importantly, 80% of the foundation's staff are women from urban slum backgrounds themselves, embodying the ethos of empowering women to become agents of change within their own communities.

The foundation's approach is deeply rooted in community engagement and input. Through ongoing collaboration, surveys, and field-based programming, Myna Mahila Foundation has cultivated a deep understanding of the needs, challenges, and aspirations of the communities it serves. This firsthand knowledge informs every aspect of the Myna Bolo solution, ensuring that it is tailored to the specific realities and contexts of urban slum life.

Furthermore, the Myna Mahila’’s track record of success in providing sexual and reproductive health services to over 1.5 million women demonstrates its expertise in addressing the unique healthcare needs of marginalized communities. Through initiatives such as the MHealth telehealth clinic program, Myna Mahila Foundation has already made significant strides in improving access to affordable healthcare and empowering women to make informed decisions about their health.

Crucially, the design and implementation of the Myna Bolo solution are guided by inputs from the communities themselves. Women from urban slums are employed through the Rani Jobs platform to generate training datasets for the AI model, ensuring that the data captures the lived experiences and realities of the target population. Community workers trained by health experts like Dr. Shraddha Kale Kapile play a vital role in fine-tuning the AI algorithms and validating the accuracy of the responses provided by the chatbot.

Through this co-creation process, Myna Mahila Foundation aims to address the gap in access to new-age technologies among low-income communities while providing contextualization in the healthcare industry. By empowering women from urban slums to become active participants in the development and implementation of the Myna Bolo solution, the foundation is fostering meaningful and sustainable change from within the communities themselves.

We are currently in the pilot phase as our proof of concept is positioned for scale after a rigorous 9 month period of development, testing, and evaluation with over 350  end users, a network of local doctors and various public health, AI, and government advisors. With over 85% medical accuracy within family planning and 90% accuracy with localisation of the responses from the chatbot, we plan to launch Myna Bolo within the urban slum communities in a multi-phase approach through the AI Assisted Healthcare Worker system.  

Key milestones include formalizing partnerships with local governments, and measuring changes in women's health seeking behaviors and sense of agency through user metrics. To further increase user engagement on Myna Bolo we will integrate Myna Bolo into Whatsapp. The focus is on validating the chatbot's effectiveness in real-world settings and refining its capabilities based on feedback and evaluating how a human in the loop provides a way for effective and long lasting impact. 

Additionally, as we shift towards scaling and growth with our project, we will solidify our business model and partners to ensure our solution is financially sustainable and can be a long term solution. We have already initiated partnerships with local government and private sector partners like Microsoft and plan to formalize these in the coming months.  

We believe MIT Solve will be a catalyst for getting our solution from a Pilot phase to a Growth phase. While Myna Mahila Foundation has made significant progress in leveraging technology to empower marginalized women and girls in urban slums, we have faced challenges in evaluating technical and impact, and we believe Solve can provide the resources and expertise to help us overcome them.

One of our main challenges lies in the technical aspects of our solution. We are seeking tailored capacity workshops around AI chatbots and scaling to enhance our understanding and capabilities in this area. Additionally, access to in-kind and pro bono resources for tools such as OpenAI credits would help facilitate the development and deployment of our technology, enabling us to reach more women and girls with accurate and timely sexual and reproductive health information.

In addition, we are also seeking assistance with monitoring and evaluation. We aim to build an impact measurement practice for our AIHS model, enabling us to assess and optimize the effectiveness of our solution in improving the sexual and reproductive health outcomes of marginalized women and girls. By receiving guidance and resources in developing evaluation frameworks and methodologies, we can better understand and communicate the impact of our work in order to attract more partners such as private healthcare providers and local government partners. 

While funding is not our sole focus in applying to Solve, we recognize its importance in overcoming financial constraints and accelerating the development of our solution. Funding from Solve would enable us to make major updates to our existing applications, as well as invest in backend verification algorithms to enhance the accuracy and reliability of our chatbot responses. Ultimately, this support would enable us to scale our impact and reach even more women and girls in need.

Through our novel feedback loop between RANI workers' input in training AI models and assisting healthcare workers holds promise, we tackle the complexity of diverse sexual and reproductive health queries influenced by offline social norms. Misconceptions like "TB treatment hindering conception" need nuanced guidance. Beliefs such as "breastfeeding from one breast causes cancer" require thorough clarification. Misunderstandings like "lack of pleasure due to a bodily issue" demand empathetic responses. If successful, this approach's effectiveness in enhancing medical responses and local relevance could extend to the legal domain, addressing issues like domestic violence and  land rights. Despite challenges, successful implementation has the potential to significantly improve women's lives while respecting social norms.

Our solution via Myna Bolo takes an innovative approach to address sexual and reproductive health (SRH) and economic mobility. We recognize the link between these two factors, and with our solution aim to provide jobs as well as access to reliable information and resources. In doing so, we are approaching SRH, a stigmatized topic, in a manner that empowers women and girls and gives them agency.  Our solution also allows for women to gain skills via their training and work and enter the technology space, remotely. This model ensures that women who want to work and upskill themselves, have the opportunity to do so. Simultaneously, through their work and training, they are gaining a better understanding of SRH.  Thus, this is a community centered, unique solution that tackles the problem from multiple angles. It would demonstrate how integrating technology to ensure access to education and resources can provide more than just an answer to a question. It helps build communities, allows for the dissemination of evidence based information and creates jobs.

We believe Myna Bolo and the AIHS model’s success could transform the sexual and reproductive health space in urban slums. After testing, scaling, and measuring its impact, we plan to  make the model open access and share it with a variety of NGOs, government organizations, and any other stakeholders looking to empower their beneficiaries with critical information pertaining to SRH, including non-judgemental personalized solutions. In the future, this localized SRH chatbot will also be tailored for rural areas, and will be expanded to support women in other domains as well, such as with information related to livelihood and career development. We could follow the same set of steps needed to tailor the application to the livelihood domain to provide women with personalized encouragement and locate local jobs available to match their needs.

Our project, centered on the Myna Bolo chatbot, is designed to bridge a critical gap in AI technology, providing employment for women from India's underserved communities and increasing sexual and reproductive health education. Current LLMs struggle to meet the unique needs of women: lacking in recognizing local dialects, cultural nuances, and behavioral patterns, especially for marginalized women. We aim to address these issues with the implementation of Myna Bolo.

Through this project we are providing women with a remote and flexible job. This job will present an opportunity for women to work and gain experience in a technical setting. It will allow them to be at the forefront of technological innovation, hopefully inspiring them to pursue further work and careers within technology. They will gain cross-dimensional skills that can be applied to a variety of settings. This job will help them gain agency and independence over their actions outside of their home-life. 

Being trained as prompt engineers and healthcare specialists will also provide women and girls access to evidence-based information regarding their sexual and reproductive health. As they are training, and serving as the human-in-the-loop aspect of AI, they will gain a better understanding of their SRH. In doing so, they will also be sharing this information with Myna Bolo users, thus helping others take care of their own health. Through this widespread information, women and girls will feel more empowered to help those around them, knowing that their source of information is reliable.Overall, the Myna Bolo Chatbot will only serve as a positive addition to the communities in which it is implemented.

Our impact goals are multifaceted, aiming to assess the effectiveness of our intervention and its impact on women's sexual and reproductive health (SRH) knowledge and decision-making in urban slum communities. One objective is to conduct comprehensive research to evaluate the impact of Myna Bolo on improving SRH knowledge and decision-making among women. Success will be measured by significant shifts in SRH knowledge, health behaviors, healthcare service utilization, and decision-making confidence, identified through pre- and post-intervention surveys and focus group discussions. 

Another objective is to analyze the effectiveness of the Human-in-the-Loop (HITL) process in enhancing user engagement and chatbot response accuracy, particularly in sensitive SRH discussions. Success will be gauged through user interaction data analysis, feedback from RANI workers and healthcare specialists, and improvements in chatbot response accuracy and relevance. Furthermore, we aim to investigate the impact of anonymity provided by the AI chatbot on women's willingness to engage in SRH discussions, especially on stigmatized topics, through a randomized controlled trial (RCT). 

We will measure effectiveness through changes in the frequency and depth of SRH queries post-chatbot introduction, compared to traditional face-to-face consultations. Additionally, we prioritize enhancing data security and developing robust models to ensure the accuracy of the Myna Bolo chatbot. Success will be evaluated by the chatbot's ability to securely handle sensitive user data and provide precise responses. Through these objectives, we seek to comprehensively evaluate the impact of our intervention and continuously improve its effectiveness in addressing SRH needs in urban slum communities.

Our impact goals align with the United Nations Sustainable Development Goals (SDGs), particularly SDG-3 (Good Health and Well-being) and SDG-5 (Gender Equality). By providing women with unbiased, factual SRH information through the Myna Bolo chatbot, we aim to address barriers to accessing healthcare information and promote gender equality. Through ongoing monitoring and evaluation, we strive to continuously improve our intervention and maximize its impact on the lives of women in underserved communities.

Our solution uses advanced AI technology to break down language and literacy barriers in healthcare communication for marginalized communities. Our main tool is the Myna Bolo chatbot, which we're enhancing to speak the language and understand the cultural nuances of these communities.

Here's how we're doing it:

1. Training Data Development: We're working with the Myna Mahila Foundation's Rani Work initiative to gather insights from local women in urban slums. They help us understand the language, beliefs, and norms related to sexual and reproductive health (SRH), which we use to train the chatbot.

2. Voice Recognition: We're improving Myna Bolo's ability to understand and respond to spoken queries in different local dialects, making it more accessible to users.

3. Testing and Improvement: We're continuously testing and refining the chatbot with feedback from diverse user groups, ensuring it addresses cultural sensitivities and provides accurate healthcare information.

4. Knowledge Base Compilation: We're building a database of relevant information tailored to the context of urban slums, ensuring the chatbot's responses are up-to-date and informative.

To bring this technology to the community, we're training local individuals to become AI-assisted healthcare workers. They'll educate women in urban slums on how to use the chatbot to access healthcare information, serving as a human connection for those who need further assistance. Through this approach, we aim to empower marginalized communities with accurate healthcare knowledge and support.

In the Myna Mahila Foundation, we have a total of 60 team members. Specifically for the Myna Bolo and AIHS project, there are 15 active team members. This includes field staff operating in urban slum centers, our founder, a program manager, a data associate, Quality Assurance Supervisors, an Accounts and Payments Executive, a product manager, and a technology vendor contracted through an external agency.

1 year, however our solution is built on 8 years of experience in urban slum settings on employment and sexual and reproductive health interventions.

Our team is deeply committed to fostering diversity, equity, and inclusion in every aspect of our work. We prioritize gender inclusivity and have designed our project to cater to diverse individual preferences and abilities, particularly focusing on women's needs. Our multilingual interfaces and user-friendly designs accommodate varying literacy levels, while features compatible with assistive technologies ensure accessibility for individuals with different levels of literacy. To further support diverse learning styles, we're developing training modules in multiple formats, including videos, audios, and interactive sessions.

Myna Mahila Foundation, has a longstanding history of working with religiously, culturally and socioeconomically diverse, and marginalized communities, particularly women living in urban slums. Our RANI workers, who are women from slum communities, have been pivotal in our efforts over the last 3 years. Many of these women have never been employed before and face significant barriers due to traditional social norms. Through our grassroots initiatives, we've reached over 1.5 million women in urban slums, providing essential hygiene products and sexual and reproductive health education and services. A significant portion of our beneficiaries come from low-income backgrounds. We recognize the importance of challenging these social norms to empower women to realize their full potential and advocate for themselves.

Our organization is predominantly field-based, with 70% of our team composed of women from urban slums who are at the core of our projects. These women not only execute plans on the ground but also provide invaluable input, drawing from their lived experiences and deep understanding of the communities we serve. We prioritize providing opportunities for growth and leadership for these women, recognizing their expertise and commitment to our mission.

Our business model revolves around providing value to our key customers and beneficiaries, primarily focusing on empowering women from urban slums through AI training and expertise.

For our key customers, which are the women from urban slums, we provide specialized training programs in healthcare and legal domains. Initially, our focus is on strengthening the expertise of AI-Assisted Healthcare Specialists (AIHS) through advanced courses and mentorship opportunities. These women are seeking to develop AI skills to improve healthcare access and quality in their communities. By offering training in healthcare and legal domains, we enable these women to become domain-specific experts, equipped with the knowledge and proficiency to provide accurate and empathetic AI-powered solutions.

We provide these training programs through a combination of in-person and online platforms, ensuring accessibility and flexibility for our beneficiaries. The courses cover a range of topics, including data pre-processing, language and content generation, AI output evaluation, and domain-specific knowledge. Through hands-on learning experiences and mentorship, we equip women with the skills and confidence to apply AI technology effectively in their respective fields.

Our beneficiaries want and need these training programs to access new opportunities for personal and professional growth. By acquiring AI expertise, they can enhance their employability, contribute to improving healthcare and legal services in their communities, and ultimately, empower themselves to drive positive change. Additionally, these programs provide a pathway for women to break free from traditional social norms and gain recognition as local experts.

In the short term, our financial sustainability plan focuses on securing grants that align with our mission of empowering underserved women with AI-driven healthcare knowledge. These grants will support the pilot phase of our project and enable us to diversify community testing. To date, we have successfully secured grants from reputable organizations such as Bill and Melinda Gates Foundation, Data.org, and The Agency Fund, providing evidence of our ability to attract funding for our work.

Looking ahead, our long-term financial sustainability strategy involves forging strategic collaborations with government agencies and prominent non-governmental organizations (NGOs). These partnerships will not only expand our reach but also create opportunities for sustainable revenue streams. We plan to introduce operational costs, including technology support, field team salaries, and program maintenance, which will be covered by charging fees for our services. 

By leveraging these partnerships and implementing a fee-based model, we aim to establish a self-sustaining financial model that covers ongoing expenses and ensures the continuity of our initiative. Additionally, we are exploring other revenue-generating opportunities, such as selling products or services, service contracts, and raising investment capital, to further diversify our funding sources and enhance our financial resilience.