Using AI-Powered Survey Agents in Real-Time Assessments of Humanitarian Response
Manifestation | En ligne
-
Organisé par:
UNICEF
À propos de l'événement
I present how a combination of traditional qualitative interviews and an AI-powered survey agent was used to support an After-Action Review (AAR) of UNICEF’s humanitarian response to Typhoon Yagi in South-East Asia. This innovative approach aimed to improve the timeliness, inclusiveness, and analytical depth of post-crisis learning by leveraging artificial intelligence alongside established evaluative practices. The session includes a live demonstration of the AI survey agent used during the review, and reflections on its value for real-time evidence generation, especially in humanitarian contexts.
Overall agenda:
1. Context/Purpose/Challenge - 3min
2. Traditional vs. AI-Enhanced Data Collection - 5min
3. Demonstration of the AI Survey Agent - 7-10min
4. Evaluation Insights - 5min
5. Implications for evaluation practice - 5min
6. Q&A and discussion - 10min
QR code will be shared for audience to test AI agent firsthand.
Overall agenda:
1. Context/Purpose/Challenge - 3min
2. Traditional vs. AI-Enhanced Data Collection - 5min
3. Demonstration of the AI Survey Agent - 7-10min
4. Evaluation Insights - 5min
5. Implications for evaluation practice - 5min
6. Q&A and discussion - 10min
QR code will be shared for audience to test AI agent firsthand.
Conférenciers
| Nom | Titre | Biography |
|---|---|---|
| Ali Safarnejad | Multi-Country Evaluation Specialist | I am a monitoring, evaluation practitioner, having worked in the field for four UN agencies in the last 2 decades. Currently I'm the multi-country evaluation specialist at UNICEF covering Thailand, Lao PDR and Vietnam. My academic background is a mix of development, engineering and science. |
Résumé
The AI survey agent deployed rapidly, achieved high response rate, and enabled flexible, multilingual, and engaging participation. The experiment shows AI can automate parts of the evaluation workflow, supporting—but not replacing—evaluation management. It holds promise for democratizing evaluation, making evaluation practice more accessible with smaller budgets. In the Typhoon Yagi AAR, it proved useful, and it may benefit both humanitarian and development evaluations with greater cost-efficiency and inclusiveness.
Further enhance the agentic survey tool with a "management interface", and real-time analysis to allow sampling for evidence convergence rather than a priori sample size.