AI in Evaluation: From Interviews to Insights
Seminario web | En línea
-
Organizado por:
UNICEF
Sobre el evento
This session presents a field experiment from an evaluation in which AI-assisted and traditional qualitative interviews are compared. It also showcases complementary innovations: an AI social media listening tool developed by UNICEF’s SBC team and an AI-based equity analysis tool from India. Together, these applications illustrate how AI can expand data sources, deepen insights, and support more adaptive, inclusive evaluation practice.
Presentador/a
| Nombre | Título | Biografía |
|---|---|---|
| Ali Safarnejad | Multi-Country Evaluation Specialist | Ali is an evaluation and data science practitioner, having worked in the field for four UN agencies in the last 2 decades. Since 2020 he has led and managed evaluations for UNICEF across 17 countries in Asia and the Pacific. His academic background is a mix of development, engineering and science. |
| Daniel Patrick Alonso Valckx | Multi-Country Evaluation Specialist | Daniel is an Evaluation Specialist at UNICEF, and formerly in Evaluation roles at UNDP and the African Development Bank. He has a Master’s degree from the École Nationale d’Administration in France, a Master’s from the College of Europe, Bruges, and he is pursuing a PhD in evaluation & development. |
| Juan Pablo Arguello | Social & Behavior Change Specialist | Juan Pablo Arguello is a Social and Behaviour Change Specialist and Project Manager with 10+ years’ experience in digital engagement, social listening, and community participation. He has led UNICEF platforms like U-Report and IoGT, advancing youth engagement, AAP, and evidence-based, people-centred digital innovation. |
| Atishay Mathur | Research & Evaluation Specialist | Atishay Mathur is a Research & Evaluation Specialist at UNICEF. He has 10+ years’ experience working in India & the UK leading evaluations in the social policy, education, protection, health, skilling, & SBC space. He is a 2025 Global Data Fellow at the University of Edinburgh's Childlight Programme |
Moderador/a
| Nombre | Título | Biografía |
|---|---|---|
| Jane Mwangi | Regional Evaluation Adviser | Jane is the UNICEF Regional Evaluation Adviser in the Asia and Pacific region. |
Resumen
The session demonstrated that AI is already a practical reality in UNICEF evaluation work, not a future aspiration. Three distinct tools — an AI conversational interviewer, a social and community listening platform, and an equity quality-assurance tool — showed AI can simultaneously expand scale, improve consistency, and deepen qualitative richness. Critically, all presenters converged on a shared principle: AI augments human judgment rather than replacing it. Ethical governance, human oversight, and inclusive design are not add-ons but preconditions for responsible deployment.
Immediate priorities span three areas. All three tools require continued testing in diverse contexts, with systematic documentation of lessons. Technically, adaptive probing logic, contextual memory, auto-coding, and smarter filtering need refinement. On inclusion, audio-visual interfaces and child-friendly designs will broaden reach. Governance must keep pace: updated AI ethics frameworks, alignment with UNICEF data protection policy, and IRB protocols for any child-facing tool. Longer term, all three tools are open to external collaboration — interested organizations are encouraged to reach out.