53110ListenFirst: Leveraging natural language processing to amplify underprivileged women’s voices for global development fundersClosedStudyInnovate UKListenFirst's project is to develop an interactive natural language processing (NLP) and text analytics platform that aggregates and analyses text from grassroots organisations' interviews with underprivileged women in South Asia, making first-hand accounts available to development finance institutions, foundations and gender smart investors in the UK and Europe who deploy billions for gender equity in emerging markets. Global aid and private funding for gender equity have rapidly increased in the last few years, but this is not making a proportionate difference to women's lives on the ground. This is because well-intentioned funders are far-removed by distance and lived experience from the vulnerable women they seek to empower, and lack the firsthand data they need to make effective interventions. At the same time, millions of scattered grassroots organisations are in continuous dialogue with vulnerable communities but lack the incentives and resources to fully utilise this unstructured data for themselves or others. Current modes of data collection are slow and resource-intensive. Each time a query arises, funders typically have to engage local intermediaries, who in turn engage grassroots organisations to collect data in person. Much is lost in translation in this time-consuming, expensive and person-dependent analysis process, and hours of recorded conversations are discarded after a single use, though they may contain many insights relevant for other purposes. It is therefore no surprise that two-thirds of the data needed to achieve Sustainable Development Goal 5 on Gender Equality is unavailable, despite estimates that a household in India is surveyed 10 times a year on average. ListenFirst bridges these gaps by aggregating rich unstructured data from grassroots organisations' recorded conversations with women onto a single digital platform, and using text analytics and NLP to derive actionable insights for global funders. While NLP and text analytics are routinely applied to better understand consumers and improve FMCG product design and marketing, these technologies are relatively unfamiliar to the development sector despite an arguably greater need for nuanced, cost-effective analysis of diverse community needs. We believe that this project can have wide-reaching social impact, by amplifying women's voices and creating a culture that values their opinions on interventions that seek to serve them. With women and girls forming 70% of the world's poorest 1.5 billion, and 80% of people displaced by climate change, their social inclusion is pivotal to move the needle on global development.