Mainstream scientific and traditional knowledge systems have the potential to restore degraded habitats, enhance biodiversity-based enterprise incomes and mitigate climate-induced disasters. However, traditional knowledge systems are at increasing risk of not being transmitted intergenerationally. On the other hand, mainstream scientific knowledge systems are rarely translated into community-relevant content. Together, these barriers to knowledge transmission hamper the pursuit of human wellbeing and biodiversity conservation locally. We have elected to use an AI-first strategy to reverse this. We are building a set of ChatGPT-powered WhatsApp chatbots (GramMitra) so local communities and other stakeholders can access scientifically curated knowledge to (a) sustain and enhance their biodiversity-based livelihoods, (b) restore agroecosystems, and (c) conserve and manage biodiversity in their surrounding landscapes. These chatbots will provide user-friendly, interactive and efficient ways to access information with a conversational interface. Users will be able to instantly address queries and obtain context-specific, summarised responses in their own languages, without having to browse multiple webpages.