Gooey.AI
Gooey.AI – a low-code platform for building and deploying multilingual AI copilots over WhatsApp, voice, SMS, and web, with proven frontline health and agriculture deployments (Farmer Chat, UN IOM Afiya) directly relevant to NAISH delivery.
Gooey.AI is a low-code AI orchestration platform for building and deploying multilingual AI "copilots" for frontline workers and underserved populations in the Global South. Teams compose AI workflows ("recipes") from LLM prompts, knowledge-base files, code, and evaluation datasets, then deploy them with one click to WhatsApp, voice/IVR, SMS, and web. It is model-agnostic (Google, OpenAI, Meta, Mistral, DeepSeek from one account) and markets support for 1000+ languages. It is included in this archive because its deployment pattern – knowledge-grounded, multilingual copilots delivered where people already are – is a proven delivery layer for NAISH health and agriculture use cases.
Reference: gooey.ai
WhatsApp-first, measurable, open. Copilots answer from a vetted knowledge base with page-level citations, support voice notes and image input for low-literacy users, and track adoption/outcomes with "golden" evaluation datasets. Core components are open source (Apache-2.0): github.com/GooeyAI.
Farmer Chat
Farmer.CHAT is a conversational AI agricultural-advisory copilot giving smallholder farmers vetted, locally relevant guidance on planting, irrigation, pest management, fertilisation, and climate risk. It is trained exclusively on trusted local resources – farmer-to-farmer videos, government extension documents, and best-practice manuals – and built by Gooey.AI with Digital Green. It is multimodal (text, audio, image) and multilingual (Hindi, Telugu, Bhojpuri, Amharic, Swahili, English), delivered over WhatsApp, Telegram, and web.
Reported outcomes include high rates of farmers applying recommendations within 30 days. (Larger program-wide figures cited on the site reflect Digital Green's overall extension reach rather than chatbot users specifically.)
- Page: gooey.ai/farmer-chat · Live agent: farmerchat-via-gpt-4o · Case study: help.gooey.ai/farmerchat
UN IOM – Afiya
Afiya is an AI health agent on WhatsApp created by the UN International Organization for Migration (IOM) office in Tunisia and powered by Gooey.AI. It serves migrants in Tunisia and neighbouring countries with health information (vaccination, pregnancy, mental health) and service navigation. WhatsApp was chosen because migrants often lack a local SIM card.
Features: 24/7 multilingual support (20+ languages incl. regional dialects), voice messaging for low-literacy users, structured care-pathway guidance, and an integrated service locator of 2,000+ GPS-tagged health facilities linked to Google Maps.
- Page: gooey.ai/un-iom · Live agent: iom-tunisia-health-chatbot
Other deployments
Medic (Kenya)
AI assistant for Kenya's community health workers (Rockefeller Workflow Accelerator, 2025).
Medtronic Labs (Bangladesh)
Bilingual (Bangla + English) WhatsApp agent for health workers managing diabetes and hypertension patients.
Mshauri (Kenya)
WhatsApp agri-advisory bot in the Agriculture Information Exchange Platform (backed by Germany's BMZ / GIZ); text, voice, and crop images in English and Swahili.
City of Seattle
Multilingual voice agent for government services (Rockefeller Accelerator cohort).
Gooey.AI's 2025 Workflow Accelerator was supported by the Rockefeller Foundation. Its agriculture-advisory work sits within the broader Gates-Foundation / GIZ / CLEAR Global ecosystem for AI-based advisory services for smallholder farmers.
Relevance to NAISH
- WhatsApp-first, multilingual delivery matches how Nigerian frontline workers and the public already communicate – the exact pattern of Afiya (Tunisia), Medtronic Labs (Bangladesh), and Medic (Kenya), extensible to Hausa, Yoruba, Igbo, and Pidgin with voice for low-literacy users.
- Knowledge-grounded, cited answers support clinical safety and trust for health advisory.
- Referral & service-locator pattern (Afiya's 2,000+ facilities, 1,707 referrals) is a transferable model for connecting patients to Nigerian primary health centres.
- Open-source, self-hostable components (Apache-2.0 server + embeddable widget) support data sovereignty and local hosting on NAISH infrastructure.
- Measurability aligns with the M&E requirements of Gates-funded programs.
Links
Masakhane
Masakhane – the grassroots pan-African NLP collective building open datasets, models, and benchmarks for African languages, including Nigerian languages and health-relevant resources directly reusable by NAISH.
Securing AI-Powered Solutions
A general guide, distilled from reviewing real AI systems, to the security controls and tool-selection choices that make an AI solution safe to deploy at national scale.