Toward Neurodivergent-Aware Productivity: A Systems and AI-Based Human-in-the-Loop Framework for ADHD-Affected Professionals Conference: arXiv; July 2025;

Raghavendra Deshmukh · 2025

Read Toward Neurodivergent-Aware Productivity: A Systems and AI-Based Human-in-the-Loop Framework for ADHD-Affected Professionals Conference: arXiv; July 2025; by Raghavendra Deshmukh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-06864.txt.

Introduction

Raghavendra Deshmukh proposes a productivity assistant designed for professionals with ADHD rather than adapted to them afterward. Knowledge work demands juggling tasks and regulating attention, and standard tools ignore how ADHD minds fluctuate. The framework senses tab use, app focus, and inactivity on the device itself. It infers the user's attention state with machine learning and responds with gentle nudges, reflective prompts, or a virtual body-double presence. Privacy is protected by keeping the sensing local.

The design blends systems thinking, where attention is shaped by feedback loops, with human-in-the-loop AI that supports rather than interrupts. Deshmukh argues low-touch help fits what these users actually accept: heavy dashboards add the overload they are meant to cure. The paper is a framework proposal, not a trial, offered as a replicable model for inclusive tools. Its test would be whether such quiet support measurably steadies a working day.

Catalog record

CREATORRaghavendra Deshmukh
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-06864

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Toward Neurodivergent-Aware Productivity: A Systems and AI-Based Human-in-the-Loop Framework for ADHD-Affected Professionals Conference: arXiv; July 2025; Raghavendra Deshmukh Abstract . Digital work environments in IT and knowledge-based sectors demand high levels of attention management, task juggling and self-regulation. For adults with Attention-Deficit/Hyperactivity Disorder (ADHD), these settings often amplify existing challenges such as time blindness, digital distraction, emotional reactivity and executive dysfunction. Such individuals prefer a low-touch, easy-to-use interventions to help them in their day-to-day activities. Conventional productivity tools fall short of supporting the cognitive variability and overload patterns experienced by neurodivergent professionals. This paper introduces a comprehensive framework that blends Systems Thinking, Human-in-the-Loop with Artificial Intelligence (AI), Machine Learning (ML) and privacy-first adaptive agents to support ADHD-affected users in managing digital work. At the heart of the solution is a voice-enabled productivity assistant that senses user behavior—tab usage, application focus, inactivity windows—using lightweight, on-device machine learning. These behavioral cues are analyzed in real-time to infer attention states and deliver adaptive nudges, reflective queries or accountability-based presence (body doubling) designed to co-regulate cognition without disruption. While technically grounded in AI/ML, the design is deeply influenced by Systems Thinking, viewing user attention as a product of dynamic feedback loops. This hybrid approach bridges behavioral sensing with cognitive inclusivity, offering a replicable model for adaptive, neurodivergent first decision support tools in high-distraction work environments. Acknowledgements. This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0). To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/ . 1. Introduction It is estimated that 4% to 5% of adults globally are affected by ADHD ( nim [n. d.] ) . Among working professionals, the number is likely underreported due to factors such as late diagnosis, masking behaviors, and ongoing stigma in corporate environments. With the rise of remote-first and digital-first work, the challenges associated with ADHD—once partially buffered by in-person environmental scaffolds—have become more visible, disruptive, and harder to manage. Digital work environments demand sustained attention, executive control, and rapid context switching, conditions that are inherently misaligned with the core difficulties experienced by ADHD-affected individuals. These include time blindness, emotional reactivity, disorganized thinking, and cognitive fatigue. Studies show that executive dysfunction—difficulty with prioritizing, organizing, and initiating tasks—impairs day-to-day workplace functioning ( Townsend 2019 ; Amador-Campos and Krieger 2023 ) , often resulting in missed deadlines, impulsive switching, or difficulty following through. The fragmented nature of knowledge work, which requires juggling between multiple tools and platforms, contributes to heightened cognitive load and task inertia ( Adler et al. 2017 ) . Additionally, many professionals with ADHD experience emotional dysregulation, leading to difficulties managing feedback, peer interactions, and stress in high-performance environments ( Shaw et al. 2014 ) . These barriers are further amplified by the rise of “hustle culture”—a mode of work that glorifies long hours, constant output, and perpetual availability. Particularly prevalent in tech startups and high-velocity digital teams, this ethos incentivizes overcommitment and hyper-productivity, which can be cognitively unsustainable for neurodivergent workers. A recent PLOS ONE study identified how non-attachmental and hustle-driven work cultures can undermine psychological stability and exacerbate attention-related distress, especially among workers with low emotional buffer capacity ( Rashid et al. 2024 ) . Similar findings from organizational studies point to hustle culture as a path to burnout, highlighting how its pressures erode attention regulation and reduce sustained focus in cognitively diverse teams ( Valentina Zhykovska [n. d.] ) . Most existing tools for ADHD-affected professionals are structured around task reminders, habit tracking, or productivity logs. Although many incorporate machine learning to detect usage patterns, they often rely on shallow context-awareness and fail to support adaptive emotional scaffolding or dynamic cognitive states ( Bhandarkar et al. 2024 ) . These tools tend to prioritize behavior correction over self-awareness and flexibility. What is missing is a holistic approach that treats ADHD not as a task deficit but as a systems-level attentional modulation challenge. Current interventions rarely take into account the feedback loops between cognition, emotion, and context. Moreover, most tools lack privacy-first architectures and operate with opaque data collection, a problematic pattern given the vulnerability and stigma around neurodivergence in workplace cultures. In response, this paper introduces a conceptual framework for an AI-powered, voice-enabled assistant designed for ADHD-affected professionals working in digital settings. The system operates fully on-device and uses lightweight behavioral sensing—such as tab churn, inactivity, and app-switching patterns—to infer attention states and deliver adaptive, soft-touch nudges. These may take the form of reflective prompts, brief accountability cues, or body doubling-inspired co-working signals, all of which are governed by user-controlled settings and local privacy rules. At its core, the assistant applies systems thinking—viewing attention as an emergent property of interaction, rather than a static user trait. A human-in-the-loop feedback layer allows the tool to evolve with the user’s preferences, maintaining relevance and personal alignment. This hybrid model prioritizes presence, flexibility, and neuroinclusive design, offering an emotionally safe alternative to conventional task-centric interventions. Finally, to ground the tool in real-world insight, we employed a co-design methodology. A preliminary survey with 25 ADHD-affected professionals revealed key struggles with distraction, motivation, and digital fatigue. Their inputs directly shaped the tool’s behavior logic and user interface. One of the authors—an ADHD-affected software developer—provided continuous “lived input,” anchoring the assistant’s design in contextual authenticity. 2. Literature Review 2.1. Using AI/ML in ADHD support: Trends, diagnosis vs assistive models There are many applications to support ADHD which have been built using AI and ML technologies and algorithms. They have been focussed on diagnosis, classification, symptom prediction. For example, studies such as those by Chen et al. (2023) and Maniruzzaman et al. (2022) ( Chen et al. 2019 ; Maniruzzaman et al. [n. d.] ) have used deep learning to distinguish ADHD vs non-ADHD individuals using clinical, behavioral or sensor data. These systems are typically built around pathology detection using large labeled datasets in controlled environments. However, few systems apply AI to everyday support for adult ADHD populations. Some emerging tools like EndeavorRx ( noa [n. d.] ) for children or productivity-focused apps like the one quoted by Bhandarkar et al. (2024) ( IEEE10910758 ) offer reminders or nudges, but remain behavior focused and lack real time adaptation. Moreover, these tools rarely account for cognitive state fluctuation, emotional reactivity or individual rhythms. There exist mobile apps that allow ADHD affected students and adults to set their regular routines and assist them manage their schedules via alerts and insights ( Knouse et al. 2022 ; Kyriakaki and Driga 2023 ) . The proposed work addresses this gap by applying lightweight, privacy-first AI not for adaptive support in professional work environments and not diagnosis. The system proposed here responds to moment-by-moment behavioral patterns such as tab usage or inactivity while preserving user autonomy and data privacy. 2.2. Productivity Tools for Neurodivergent Users – Challenges with One-Size-Fits-All Design A wide range of productivity tools such as Trello, Todoist, Notion, Calendars etc are widely adopted in digital workspaces to help users manage tasks, deadlines and goals. While these systems offer powerful scheduling and collaboration capabilities, they often operate on the assumption of sequential planning, sustained attention and consistent executive functioning. However, for neurodivergent users, especially those with ADHD, such tools can become sources of overwhelm rather than support.

Read the complete text