Skip to content
ESAPIENS Solutions
All projects

What 4,800 participants said about a regional leadership program

AI-assisted analysis of open-ended feedback in three languages to decide what to keep, fix and scale in the next cohort.

Laptop showing survey charts on a desk

The challenge

After three cohorts of a leadership program for public officials, the team had close to 4,800 open-ended answers in English, Spanish and Portuguese. Nobody had read them all, and decisions about the next edition were being made on satisfaction scores alone.

Our approach

We used language models to group every answer into themes in its original language, then two evaluators reviewed each theme, merged overlaps and checked a sample of answers by hand.

We compared themes by country, cohort and role, and kept a set of representative quotes for each finding so the program team could hear participants in their own words.

Results

Twelve themes explained most of the feedback. The two most frequent, mentoring time and examples from participants' own countries, became the main changes for the next cohort.

The analysis took three weeks instead of the four months a manual review was estimated to need, and the report was shared with all participating institutions.

Project details

Client
Regional development bank
Year
2025
Location
Latin America and the Caribbean

More projects

Let's talk about your next program

ESS offers knowledge solutions to support your organization's upskilling and reskilling efforts, boost capacity development and help achieve the Sustainable Development Goals.