Skip to content
Abhimanyu Gupta

Research

AI, pedagogy, and the future of learning

My work asks how artificial intelligence and educational technology can increase classroom engagement, strengthen student learning, and preserve meaningful assessment in an AI-enabled world.

01

Classroom Engagement

How technology, behavioral design, and thoughtful classroom systems can raise meaningful participation — not just attendance.

02

Student Learning

Adaptive practice, simulation, experiential work, and learning analytics that help students apply concepts to real decisions.

03

Academic Integrity

When generative AI can write the assignment, verification of genuine understanding has to be designed into assessment itself.

Method

The applications are not side projects. They are the instruments.

Design-based research: intervene in a real classroom, observe what breaks, revise. The tool is how the question gets tested.

  1. 1

    Educational problem

    Something visibly not working in a live course.

  2. 2

    Research question

    Framed so an answer could change practice.

  3. 3

    Prototype

    Built small enough to be wrong cheaply.

  4. 4

    Classroom implementation

    Deployed with real students and real stakes.

  5. 5

    Evidence & feedback

    What students did, not what they said they would do.

  6. 6

    Iteration

    Revise, discard, or generalize.

What I am trying to change

Practices worth reconsidering

Not because education is failing, but because several of its defaults were designed for conditions that no longer hold.

  • Passive lecture halls

    Participation designed into the session

  • The same four students answering

    Channels that lower the cost of speaking

  • Feedback three weeks later

    Feedback in three minutes

  • One-size-fits-all practice

    Adaptive practice tied to what a student missed

  • Assessment built for a pre-AI world

    Assessment that evidences individual reasoning

  • Policing AI use

    Redesigning the task so outsourcing defeats the point

Peer-reviewed work

My doctoral research at Ghent University examined how conceptual models can address requirements engineering problems in agile software development — the same instinct applied to a different domain: find where practice breaks down, then build something that fixes it.

  • Gupta, A., Poels, G., & Bera, P. (2023). Generating multiple conceptual models from behavior-driven development scenarios. Data & Knowledge Engineering, 145.
  • Gupta, A., Poels, G., & Bera, P. (2022). Using conceptual models in agile software development. IEEE Access, 10, 119745–119766.
  • Eight international conference papers (ER, REFSQ, SIGSAND, MWAIS and related venues), 2017–2019.

Next

Every question above has something built against it. See the experiments.

See what I build