Practical notes on AI, surveys, official statistics, and development data.

The open-source International Classifications MCP helps AI assistants discover, compare, validate, map and export official classifications and questionnaire response codelists.

The open-source Development Indicators MCP gives AI assistants one consistent way to find, retrieve and compare official statistics across major international data providers.

Impact Engines and the Samoa Bureau of Statistics tested a human-in-the-loop AI workflow to clean the Samoa Agriculture Survey 2025 microdata. The case study covers the workflow, quality control, privacy safeguards, costs, time savings, and lessons for official statistics.

AI agents can reason across complex survey workflows, but raw API access is not enough. A dedicated MCP server gives them a safer, structured bridge into Survey Solutions, combining clear tools, permission controls, and expert operating patterns.

A practical look at how a human-in-the-loop AI agent can review survey cases, detect inconsistencies, prepare cleaning logs and callback sheets, propose recodes, and turn approved decisions into reproducible Stata code.

AI agents can strengthen survey quality control by investigating anomalies, connecting evidence across cases and field teams, and producing clear findings for human review—not simply another dashboard or generic chatbot.

What locally operated AI agents could offer international development organisations: persistent project work, controlled access to tools and data, greater privacy, and practical support for teams with limited resources.

AI can accelerate digital questionnaire programming, explain validation logic, and catch structural errors, but it still needs platform-specific context and expert review. This article explores where it helps and where it falls short.

A $150 box of apparently broken computer parts became a working PC with help from AI vision, structured troubleshooting, and a little motherboard surgery—a practical example of AI augmenting hands-on problem solving.

A bring-your-own-key approach can give non-profit teams affordable access to capable AI models while improving organisational control over privacy, model selection, usage, and costs.

General-purpose AI fails in specialized environments like Survey Solutions, where "standard" code breaks under unique architectural rules. We built a custom AI assistant that combines expert-tuned logic with PDF and vision capabilities to turn practically useless generic advice into precise, field-ready solutions for global data collection.