Three use-cases of Cogniti’s educator-controlled AI agents

Cogniti, developed at the University of Sydney and piloted at several other universities, is an educator-controlled AI agent platform that is uniquely positioned to address discipline-specific AI challenges by offering tailored guidance, resources, and mentorship. This blog post explores three speculative applications of Cogniti in the humanities and social sciences, with a particular focus on its role as a qualitative research methodology coach for students. By leveraging AI-driven agents, universities and educators can create personalised learning environments that enhance research skills.

Based on Cogniti’s capabilities as an educator-controlled AI agent platform, here are three speculative use cases for humanities and social science disciplines:

1. Qualitative research methodology coach for students

Application: Create a specialised Cogniti agent loaded with qualitative research methodology resources (grounded theory texts, interview protocols, coding frameworks) that acts as a 24/7 research mentor.

Specific Implementation:

• Upload key methodology texts as resources (i.e. key authors on methodology)
• Program the agent with Socratic questioning prompts to guide students through research decisions rather than providing direct answers
• Configure it to help students work through coding dilemmas, sampling decisions, and analytical frameworks
• Enable students to upload excerpts of their data for feedback on coding approaches (with appropriate privacy safeguards)

Value: Addresses the challenge that qualitative research supervision is time-intensive and requires frequent iterations, while many students work in isolation between classes/supervisor meetings. The agent ensures consistent methodological guidance while maintaining pedagogical best practices.

2. Primary source analysis tutor for history students

Application: Deploy subject-specific Cogniti agents loaded with historical documents, secondary sources, and analytical frameworks to guide students through primary source analysis.

Specific Implementation:

• Load digitised primary sources (letters, government documents, newspaper articles) relevant to specific historical periods or themes
• Include historiographical essays and analytical frameworks as resources
• Program the agent to guide students through contextual analysis, source criticism, and evidence evaluation
• Configure it to ask probing questions about bias, perspective, and historical significance rather than providing interpretations
• Enable comparative analysis by helping students identify patterns across multiple sources

Value: Provides scalable support for developing critical historical thinking skills, particularly valuable given the labour-intensive nature of teaching primary source analysis. Students receive immediate feedback on their analytical approaches while developing critical historical insight + AI literacy.

3. Ethnographic field note analysis assistant for anthropology

Application: Create a Cogniti agent that supports anthropology students in analysing field notes and ethnographic data using established analytical frameworks.

Specific Implementation:

• Upload seminar ethnographic texts and analytical frameworks (thick description, cultural interpretation methods)
• Load anonymised exemplars of well-analysed field notes as reference materials
• Program the agent to guide students through the process of identifying cultural patterns, themes, and significance in their observations
• Configure it to prompt students to consider reflexivity, positionality, and cultural bias in their interpretations
• Enable iterative dialogue where students can refine their analytical approaches through guided questioning

Value: Addresses the challenge that ethnographic analysis requires extensive mentorship and iterative feedback, which is resource-intensive for faculty. The agent provides consistent analytical guidance while encouraging the deep interpretive thinking essential to anthropological inquiry.

Each use case leverages Cogniti’s core strengths: resource integration for domain-specific content, pedagogical steering to maintain educational value, 24/7 availability for iterative learning processes, and institutional control to ensure appropriate academic standards.

AI declaration

This blog post contains content generated and assisted by artificial intelligence (AI), specifically using Claude Sonnet 4.0 Thinking. Throughout the drafting process, Claude Sonnet 4.0 Thinking was employed to help ideate, structure, and refine key arguments and explanations. While the AI provided suggestions and synthesised information to enhance clarity and coherence, all material was carefully reviewed and edited by the author to ensure accuracy, appropriateness, and alignment with the intended message. Readers should be aware that some sections may reflect the style or phrasing typical of Claude Sonnet 4.0 Thinking’s outputs, but the final responsibility for the content rests with the author.