Five stages, each with the steps, the place of AI, and the tools that fit.
Framing the questions, the theory of change, the logframe, the indicators, and the proposal that wins the work. AI is strong at structure and weak at promises, so you supply the logic and the commitments.
Surveys, forms, mobile data collection and field logistics. AI helps you design and translate the tools; the collection itself is human and physical, and the data quality is won or lost in the field.
Cleaning, joining and analysing quantitative and qualitative data. AI accelerates the mechanical work and explains the numbers, but it must never be the final word on a figure that will travel to a donor.
Turning findings into a report, a briefing, a slide deck, or a community-facing summary. This is where AI is fastest, and where verification matters most, because a confident wrong sentence can undo months of careful work.
Making sense of what was learned, tracking recommendations, and feeding the next cycle. AI is useful for summarising and searching what you already know; the decision to change course remains human.