Practical Application of AI in Evaluation: Value, Conditions, and Strategic Integration
Classe de maître | En ligne
-
Organisé par:
Community of Evaluators South Asia
- In partnership with: Knowles Advisory and Consulting Private Limited
À propos de l'événement
AI is no longer on the horizon for evaluation. It's already here — and this 60-minute session at gLOCAL Evaluation Week 2026 will show exactly what that looks like in practice. From NLP coding thousands of transcripts in days, to predictive ML flagging, to WhatsApp chatbots reaching the most remote communities — the cost, time, and distance barriers that once defined what evaluation could do are falling away. This session draws on four real-world documented approaches and translates those cross-sector lessons into something actionable: one application to explore, one safeguard to commit to, and a clear framework for shaping AI's role in evaluation before others shape it for you. If you work in evaluation — as a commissioner, practitioner, policymaker, or alongside affected communities — this one's for you.
Conférenciers
| Nom | Titre | Biography |
|---|---|---|
| Tathagata Dasgupta | Director, Knowles Advisory | Tathagata is an experienced leader in the research industry. He is driven by data, strategy, innovation, results, and capacity building. His 30-year professional career includes work in research, MEL, and capacity building. He has considerable experience in AI, MEL, and Research Techniques. |
Résumé
This virtual masterclass on the practical application of AI in evaluation examined how AI is reshaping the MEL lifecycle, including indicator development, qualitative coding, predictive analysis, reporting, and knowledge support. Live demonstrations covered NLP-based thematic analysis with Claude, predictive modelling with Julius, and chatbot development with Nopay. Case examples from Lilapa AI, Educate Girls, and Digital Green illustrated AI use in language inclusion, educational targeting, and agricultural advisory. The session emphasised AI as a co-pilot, not autopilot, requiring human oversight, validation, transparency, and safeguards against bias, hallucination, exclusion, opacity, and de-skilling.
1) Share the session recording and presentation materials with all COE-SA members and registered participants.
2)Organise a follow-up session with Dr. Dasgupta/Knowles on practical AI tools for evaluation, grassroots activism, and community-based research.
3) Develop a short resource note on AI applications, risks, and safeguards for evaluators in South Asia.
4) Encourage participants to explore Knowles Advisory and Consulting’s website and LinkedIn page for further resources, courses, and learning opportunities.
5) Identify possible use cases where COE-SA members can pilot AI-assisted evaluation approaches with strong human oversight, ethics, inclusion, and validation protocols.
Image Gallery
