Grassroots environmental NGOs in India hold years of field data on forests, water and commons, but little of it reaches the people who could act on it. Tarkam's open-source Data Maturity Stack uses small, locally run models to help them organise, analyse, share and tell stories with their data, on their terms.
India's ecological commons (forests, grasslands, wetlands and coastal lands) sustain more than 350 million people and are deteriorating at 4% annually. The grassroots organisations restoring them hold field data no one else has: well-water levels, species records, village resource maps, oral histories. Most of it sits in notebooks, WhatsApp groups and spreadsheets that are hard for others to use, including the communities it describes.
The Stack supports the 89 organisations in the network, the rural communities they work with, other NGOs, researchers and government departments in grassroots action, evidence-based advocacy and collective stewardship of endangered ecological commons. It also offers a replicable model for community-governed digital public infrastructure.
The Data Maturity Assessment (DMA) is a guided, multilingual self-assessment of how an organisation plans, collects, analyses, stores and shares data. It results in a phased roadmap of practical next steps.
Shared Vocabulary makes data from different organisations interoperable. An NGO uploads spreadsheets or survey forms, and a small language model drafts a data dictionary (what each column means, its units and source) and links variables into a knowledge graph.
Because data on natural commons is fragmented, lacks a common vocabulary and follows mismatched administrative and natural resource boundaries, the Stack supports annotation, dictionary extraction and the creation of knowledge graphs.
The Federated Repository makes data easy to share and discover. It is open to anyone and can be searched in plain language. Grassroots organisations across the network are co-designing it, and its features, including federated architecture, natural-language querying, multilingual support, audio-video data compatibility and adherence to FAIR principles, all emerged from this process.
The Federated Repository is being co-built as an open resource with 18 organisations working in technology, research and data science, alongside Tarkam's network of grassroots environmental NGOs.
The Data Story Engine is a repository of visual assets (icons, illustrations, modified charts) and story templates. It is used to build accessible data-driven stories in local languages for diverse audiences, including government and funders.
Accessible visual language components and templates are being developed as part of the Stack to enable data-driven storytelling, designed for the rural and marginalised communities that grassroots NGOs work with.