Immunization intelligence at the last mile: Teeka seekho as a governance-first ai digital learning platform for frontline public health

Background: Digital health can extend immunization workforce learning, but many platforms are disconnected from field workflows, local-language needs, supervision, and governance of AI outputs. Teeka Seekho, the local brand for the AI-assisted Digital Learning Platform for Immunization (AI-DLPI), was developed as a regulatory-by-design public health technology for frontline immunization workers in Pakistan.

Methods: We conducted a mixed-methods, human-centred design and early-stage implementation process in Sindh, Pakistan. More than 80 stakeholders participated in the main co-design workshop series, with broader co-design rounds engaging 109 contributors across frontline, supervisory, programme, academic, and technical roles. Design inputs were translated into an Android-first mobile app, a supervisory web portal, an administrative layer, and a bounded AI assistant grounded in approved immunization content. Governance was provided through a Steering Committee and Technical Advisory Group to oversee content validation, data governance, AI guardrails, localization, and change control. The planned pilot protocol combines system telemetry, course assessments, support logs, chatbot audit logs, usability observation, surveys, qualitative interviews, and structured AI-output review.

Results: The design process shifted the platform from a conventional learning management system to a governed service ecosystem combining microlearning, searchable reference content, adaptive assessments, progress analytics, help/support functions, supervisory dashboards, and AI-assisted field support. The chatbot architecture uses backend-controlled retrieval from approved bilingual knowledge, defined response modes, out-of-scope refusal, audit logs, escalation pathways, and human-reviewed updates rather than uncontrolled autonomous learning. Early perception testing among six frontline users produced 102 marked questionnaire responses, with 75 (73.5%) marked Agree and 8 (7.8%) Strongly Agree. Participants perceived the platform as useful for vaccination work and communication confidence, but highlighted implementation priorities including offline or low-connectivity access, simpler Urdu, field scenarios, visual refinement, technical support, and strengthened chatbot accuracy.

Conclusion: Teeka Seekho demonstrates how public health technology can integrate learning, supervision, and safe AI support into one accountable workforce platform. For the Public Health and Technology session, the main contribution is not a standalone chatbot, but a transferable strategy for building AI-enabled digital health tools with fairness, explainability, safety, auditability, and implementation readiness embedded before scale-up. The next stage should evaluate usability, engagement, learning progression, supervisory uptake, grounded-answer accuracy, escalation performance, and traceable change control during routine immunization implementation.

Dr. Zahid Ali Memon is an Associate Professor & Section Head, Health Policy and Management at the Community Health Sciences Department at Aga Khan University Karachi, Pakistan. He is an interdisciplinary researcher with an impressive track record of over 20 years in reproductive, maternal, and child health research, policy, and practice in Pakistan. In his previous role, through the British Council’s Foreign, Commonwealth, and Development Office (FCDO), funded evidence for policy action for maternal and newborn health projects he contributed to bringing policy and practice changes in Pakistan. The fund supported 54 research and advocacy projects in 104 districts across Pakistan, resulting in 29 documented evidence-based policy and practice changes in the country on health information systems, service delivery, innovative community-level interventions, and health financing solutions. His further work spans health policy, implementation science, immunization systems, digital health, and public health workforce strengthening. For Teeka Seekho/AI-DLPI, he serves as the presenting author and corresponding academic contact, supporting development and pilot evaluation of a governed AI-assisted learning and supervisory platform for frontline immunization workers in Pakistan. His work emphasizes evidence generation, human-centred design, governance, and safe integration of AI into public health programmes.