The researcher.
The research.

Arunamirtharaj Chandrasekar has spent two decades in enterprise technology - managing Azure cloud accounts across Asia Pacific, working with large organisations on digital transformation, and watching AI systems being deployed at scale without the interpretive frameworks to make them useful.

The question that drove this research was not theoretical. It came from watching AI products optimise for metrics that looked right on a dashboard but produced experiences that felt wrong for users. Clicks, completions, engagement rates - all moving in the right direction. User understanding - nowhere in the model.

The ABSD framework is the answer to a question that took years to formulate clearly. Why do AI systems that see everything still miss so much? The research traces that question back to its structural cause - a gap between observable data and psychological reality - and proposes a four-layer architecture for closing it.

73% of companies cannot use their behavioural data for anything beyond basic segmentation. That number has appeared in research reports for several years. It has not moved significantly. The problem is not a lack of data infrastructure. Every serious product team has the data. The problem is a lack of interpretive framework - a principled way to move from what users do to what they need.

Four research traditions - Self-Determination Theory, affective computing, JITAI research, and learning analytics - each address part of this problem. None of them address it together. The literature synthesis for this study found minimal cross-citation between these fields. Researchers in each tradition discover similar phenomena using different vocabulary, without building on adjacent work.

ABSD is the synthesis those traditions were pointing toward without knowing it. The framework does not introduce new empirical findings. It connects existing ones into a coherent architecture and makes that architecture usable for product teams.

The timing is specific. AI now mediates the majority of customer and learner interactions across most major platforms. Gartner estimates that chatbots handle 85% of customer interactions at leading companies. The quality of those interactions depends directly on the depth of the system's user understanding. That understanding currently stops at Layer 1.

Two mentors provided guidance, challenge, and practical grounding throughout the research process.

Academic research is typically published after completion - reviewed, formatted, and released in a form that most practitioners never encounter. This research was built in the open from the start, for a specific reason.

The ABSD framework's primary claim is that practitioners already recognise the signal gap it describes - they have been working around it without vocabulary to name it. Validating that claim requires reaching practitioners, not journal readers. That requires a different kind of research presence.

This research is submitted as the Applied Research Project for the Executive MBA programme at SP Jain School of Global Management. The Applied Research Project requires original research that addresses a real business or management problem using appropriate qualitative or quantitative methods.

The ABSD framework addresses a structural gap in how AI systems understand human behaviour - a problem with direct commercial consequences for organisations building AI-mediated products. The research methodology combines expert interviews using a semi-structured protocol with a practitioner survey, analysed through thematic analysis following Braun and Clarke (2006).

For questions about the research, interview participation, or academic collaboration - use the contact form or email directly at arunchandru@outlook.com.