Mapping a GenAI landscape, and rebuilding how the research got made
An emerging-technology research practice at a global data and analytics firm, producing an insight series on generative AI, synthetic data, and adjacent startup ecosystems.
Client details anonymized; figures illustrative where confidential.
Leadership wanted credible, market-ready insight on a frontier moving faster than traditional research cycles could keep up with. The tension was coverage and depth against speed, with a small team.
I led the research, scoping trends, mapping startup product landscapes, and coordinating across internal data teams, then attacked the speed problem at its source by redesigning how the repeatable parts of each report were produced.
I designed AI-assisted workflows for the repeatable sections of each report and optimized discovery prompts against the firm's archive, improving how reliably the internal AI surfaced the right evidence.
Per-report turnaround dropped by about 20%, freeing analyst capacity for the higher-judgment work, and retrieval on the internal AI platform improved. This is where the method that now runs through all my engagements took shape.