This community innovation has been accepted at the 2026 DHIS2 Annual Conference and will be in abstract track/lightning talk.
SnapiForm: Practical AI to Eliminate Paper-to-DHIS
Across many low- and middle-income countries, routine health reporting still depends on paper forms that must be manually transcribed into DHIS2. In the Democratic Republic of the Congo (DRC) alone, this translates into hundreds of thousands of pages and millions of data points entered every month, leading to delayed reporting, staff fatigue, and persistent data quality issues. SnapiForm, developed by PATH, is a practical, needs-driven AI solution designed to eliminate this transcription burden. Built as a lightweight mini-app inside Telegram, SnapiForm enables health workers and data managers to digitize existing paper HMIS forms by simply taking a photo. Using computer vision and AI-based table and handwriting recognition, SnapiForm extracts structured data and maps it to DHIS2 data elements and category combinations. Users review and validate the data before submission using their existing credentials. Every step is logged, with an audit trail linking submitted data to the original image. A Phase 0 pilot conducted with the HMIS department from July–September 2025 in DRC demonstrated strong results. Compared to routine manual data entry, SnapiForm achieved up to a three-fold increase in data accuracy, doubled to tripled data completeness, and reduced time spent on data entry by approximately 80%. The pilot also surfaced DHIS2 metadata and configuration issues, highlighting SnapiForm’s value as a diagnostic tool for strengthening HMIS systems. SnapiForm (https://snapiform.com) is designed to be scalable, affordable, and interoperable, with current processing costs around USD 0.02 per page. Future versions will introduce offline support, multilingual parsing, and expanded interoperability beyond DHIS2. By combining human validation with transparent, modular AI, SnapiForm demonstrates how practical innovation can strengthen DHIS2 workflows without disrupting existing reporting practices.
Primary Author: Belendia Serda
Keywords:
AI, Health data quality, Paper-to-digital transformation, Computer vision
3 Likes