From Pilots to Policy: How One Innovator Wants India to Mainstream Assistive Tech
By Arunima Rajan
In an interview with Arunima Rajan, Prateek Madhav, CEO and Co-founder of AssisTech Foundation (ATF) says India’s assistive tech ecosystem is ready to scale but needs policy muscle, predictable funding and homegrown R&D to move from promising pilots to real-world inclusion.
You’ve written that AI has become a “great equaliser” for persons with disabilities. But in India, accessibility often stops at pilot projects or urban islands. What’s stopping large-scale deployment, funding, procurement processes, or policy gaps?
Prateek Madhav, CEO and Co-founder, AssisTech Foundation (ATF)
AI has immense potential to be a great equaliser for Persons with Disabilities (PwDs). However, in India, accessibility often remains confined to pilot projects or isolated urban initiatives, unable to achieve large-scale impact. From ATF’s experience of building an ecosystem for Assistive Technology (AT), supporting 500+ startups and impacting over 1 million PwDs, we have identified some key roadblocks.
Most AT innovations begin with strong intent and creativity but struggle to secure sustained capital for scaling. Traditional investors often see assistive technologies as “niche” markets with uncertain returns. At ATF, we address this through initiatives like cATalyst, our acceleration program that helps AT startups strengthen business models, access mentorship, and connect with global investors to scale beyond prototypes.
Government tendering processes typically prioritise the lowest bid and established vendors, inadvertently excluding early-stage innovators. There’s also limited provision for outcome-based procurement where success is measured by real-world impact rather than product cost which discourages adoption of AI-driven assistive tools that may initially cost more but deliver greater long-term benefits. Encouragingly, some states are now setting new benchmarks: the Governments of Telangana and Tamil Nadu have introduced direct procurement policies for assistive technologies, enabling faster adoption and creating structured demand. Such state-led championing of inclusion through AT is a welcome step toward mainstreaming accessibility.
Although India’s National Strategy for AI and flagship accessibility initiatives recognise inclusive technology, implementation remains inconsistent. Standardisation, interoperability frameworks, and dedicated funding for maintenance or capacity-building are still evolving.
At ATF, we address these systemic issues through our 4As of Framework — Acceleration, Availability, Awareness, and Association. We have seen that awareness and cross-sector collaboration are often the missing links preventing promising AI tools from reaching the people who need them most. To unlock scale, India must move from pilots to sustainable ecosystems backed by predictable funding, reformed procurement, and clear policy pathways. Only then can AI truly serve as the great equaliser it promises to be.
AI-powered wheelchairs, smart glasses, or neuro-assistive devices are often prohibitively expensive. How do you see India bridging the affordability gap, through local manufacturing, open-source models, or CSR-led partnerships?
Affordability is one of the biggest challenges in the AT space particularly for AI-powered devices that combine advanced hardware and software. Bridging this gap in India demands a multi-layered approach that balances innovation with inclusivity through local manufacturing, open-source development, and blended financing models. Building locally can significantly reduce import costs and make products more accessible. By leveraging India’s strong maker ecosystem, component supply chains, and talent in embedded systems, AT startups can develop high-quality yet affordable solutions. However, local manufacturing must go hand-in-hand with quality assurance and repair networks to ensure long-term durability and user trust.
Adopting open-source or hybrid software architectures helps lower R&D and licensing costs. Shared AI frameworks, such as speech or vision models, allow startups to customise solutions for Indian contexts without reinventing the wheel. Interoperability standards further make it easier for different devices and systems to work together, ensuring sustainable scalability.
Corporate and philanthropic partnerships play a critical role in early-stage pilots and last-mile access. Through our illuminATe market access programme, ATF connects startups directly with marginalised PwD communities and CSR partners who fund initial deployments and awareness drives. This creates a bridge between innovation and real-world impact.
For true affordability, India must institutionalise AT within public systems including subsidies, rehabilitation budgets, and pooled procurement. When AI-enabled AT is included in state reimbursement schemes, demand grows, production scales up, and unit costs decline.
In essence, affordability cannot rely on a single lever. It requires an ecosystem approach combining local innovation, open technology, CSR partnerships, and supportive policy mechanisms to ensure that AI-powered assistive tools are not a privilege, but an accessible necessity for millions of PwDs across India.
You’ve called inclusive innovation a collective responsibility of governments, investors, and communities. Are any State or Central schemes genuinely integrating AI-based assistive technologies today, or is the public system still dependent on imported or legacy tools?
India has taken encouraging steps toward integrating AI-based Assistive Technologies (AT) into public policy. National programmes like Accessible India (Sugamya Bharat) and NITI Aayog’s National Strategy for “AI for All” reflect a clear intent to promote inclusion through technology. The Ministry of Social Justice and Empowerment, in collaboration with ALIMCO (Artificial Limbs Manufacturing Corporation of India), also hosted the national conference ‘AI for Empowering PwDs’ in Bengaluru earlier this year, a strong signal of the government’s commitment to harnessing AI for accessibility.
State rehabilitation centres and government departments have begun trialling AI-powered speech aids, vision enhancement tools, and smart mobility solutions but these pilots often do not transition into sustained, budgeted programmes within schools, hospitals, or public employment schemes.
For true impact, the next phase must focus on implementation and institutionalisation. Governments need to include AI-enabled AT in their approved device lists, allocate recurring budgets for deployment and maintenance, and embed these technologies into core service delivery systems.
At ATF, we believe this is where ecosystem enablers can play a catalytic role — connecting innovators with policymakers to bridge the gap between pilot success and systemic adoption. India has already set the right intent; now, it’s time to translate that momentum into measurable, large-scale impact that reaches millions of PwDs nationwide.
AI systems are only as inclusive as the data they are trained on. Given the lack of India-specific datasets on disabilities, how are innovators ensuring that algorithms don’t reinforce bias, for instance, failing to recognise Indian languages, accents, or contexts?
The lack of India-specific datasets on disability remains one of the most critical challenges in building inclusive AI systems. Without culturally, linguistically, and contextually relevant data, algorithms risk excluding or misinterpreting large segments of Persons with Disabilities (PwDs), especially in a country as diverse as India.
To counter this, innovators are adopting multi-pronged strategies to make their models more inclusive and representative. These include curating India-centric datasets, using human-in-the-loop validation, implementing fairness and bias metrics, and developing multilingual, context-aware models that reflect India’s diversity in language, accent, and lived experience.
At ATF, we place strong emphasis on co-creation with the disability community. Startups in our ecosystem are encouraged and often mentored to involve PwDs from the earliest stages of product development. By collaborating with community organisations and user groups, they collect diverse datasets, test real-world performance, and refine algorithms based on user feedback. This participatory approach helps surface biases early and ensures that solutions are not only technically sound but also socially and culturally relevant.
Still, building ethical and trustworthy AI requires transparency, oversight, and third-party audits to evaluate fairness across disability types, regions, and socio-economic groups. The goal must go beyond technological accuracy to ensure dignity, representation, and equal access for every user.
By embedding these values into design and data practices, innovators can help ensure that AI systems in India truly understand and empower the people they are meant to serve.
You mention tools like Voiceitt and Vanderbilt’s Planning Assistant improving communication and organisation for persons with disabilities. Have you seen these or similar technologies being adopted in Indian workplaces or classrooms? What structural changes are needed to make that possible?
Adoption of advanced AI assistive tools in Indian workplaces and classrooms is nascent but growing. Private organisations and a few progressive educational institutions have piloted AAC, speech-to-text, and scheduling assistants. To scale adoption, structural changes like integrate assistive tech into formal procurement lists for schools and workplaces; invest in training for teachers, occupational therapists and HR teams so they can configure and sustain devices; and introduce accessibility audits with actionable KPIs (e.g., percentage of classrooms using assistive tools).
Digital accessibility standards must be enforced so assistive software interoperates with existing LMS and HR tools. Finally, funding models (school budgets, CSR, government subsidies) should earmark funds for assistive software licenses, device maintenance, and inclusive design consulting so tools move beyond pilots into routine practice.
At ATF, we are already working to make this vision real. Through our illuminATe programme, launched in Pune in collaboration with Bajaj Finserv, we provided AT products and services to over 3,000 children with disabilities across schools. The initiative was designed with a sustainable, scalable model, enabling future expansion with minimal external intervention.
Such partnerships demonstrate what’s possible when innovation, policy, and community action converge paving the way for India’s classrooms and workplaces to truly embrace inclusive, AI-enabled learning and employment
AssisTech Foundation has been a key enabler for startups in this space. What patterns are you seeing among Indian assistive-tech innovators, are they moving towards AI-led solutions, or are basic accessibility gaps still taking priority?
At ATF, we are witnessing a fascinating evolution in India’s AT startup ecosystem — a space that now balances immediate accessibility needs with cutting-edge AI innovation.
On one end of the spectrum, many Indian startups are still focused on addressing foundational accessibility challenges such as low-cost mobility aids, affordable prosthetics, simpler screen-readers, and basic accessible infrastructure. These solutions are crucial because they meet the urgent needs of millions of PwDs who still lack access to even the most essential assistive tools.
At the same time, an increasingly vibrant cohort of innovators is developing AI-led solutions that push the boundaries of inclusion. We are seeing startups use AI for speech-to-text recognition of non-standard Indian speech, computer vision for navigation and object identification, and smart scheduling assistants for cognitive and executive-function support.
Through ATF’s cATalyst acceleration program, we have seen a significant surge in interest among ventures integrating AI modules into their products. A promising trend emerging is that of hybrid innovation pairing inexpensive hardware with cloud or on-device AI to keep costs manageable while offering advanced functionality.
The focus is also shifting from prototypes to deployment and sustainability. Funders and accelerators are now pushing startups to demonstrate real-world viability — including user training, maintenance, and integration with public systems rather than remaining in the pilot phase.
Globally, the AI-powered Assistive Technology market is projected to grow from $22 billion in 2022 to $31 billion by 2030, with AI-based rehabilitation tools expected to triple from $2 billion to $5.6 billion between 2023 and 2031. To ensure India leads this transformation, ATF is establishing a Centre of Excellence for AI for Disability Inclusion, aimed at accelerating indigenous R&D, knowledge exchange, and cross-sector collaboration.
Many large Indian companies celebrate diversity in PR campaigns but rarely invest in accessible design. Should accessibility compliance, digital or physical, be made mandatory under CSR or workplace laws?
Yes. Making accessibility compliance mandatory under CSR or workplace statutes would accelerate inclusion, but policy design must be careful to avoid box-ticking. Embedding accessibility obligations into CSR rules (e.g., a percent of CSR spend on disability inclusion, accessible procurement requirements) could mobilise private capital for AT and inclusive infrastructure.
Similarly, strengthening workplace laws to require digital and physical accessibility (reasonable accommodation mandates, accessibility KPIs, mandatory captioning for meetings) would shift corporate behaviour from PR to practice. However, mandates should be paired with supportive measures: technical guidance for compliance, phased timelines, and third-party certification to avoid superficial compliance.
Enforcement must be outcome-oriented by measuring real accessibility outcomes (employee retention, audit scores, accessibility of digital assets) rather than just paperwork. Finally, CSR can also fund training, accessible hiring pipelines, and AT trials in smaller enterprises; pairing mandatory requirements with capacity building and incentives will yield sustainable change.
You’ve cited global projections of the AI-powered assistive technology market. But in India, how do we measure real impact, the number of PwDs who actually gain employment, education, or independence because of these tools?
At ATF, we believe that true impact is measured not by how many devices are distributed, but by how many lives are meaningfully transformed. While the global Assistive Technology (AT) market is growing rapidly, India’s success must be defined through outcome-based metrics that reflect real empowerment: in education, livelihood, and independent living.
In our work at ATF, we track both reach and results. Over the last five years, our ecosystem has impacted more than 1 million Persons with Disabilities (PwDs) through 500+ startups. But beyond reach, we evaluate how innovation translates into tangible outcomes — whether it enables employment, improves academic performance, or enhances day-to-day independence. ATF further promotes data transparency and accountability by encouraging startups and partners to use open dashboards and reporting systems, enabling policymakers and funders to identify scalable, high-impact models and avoid duplicative efforts.
Ultimately, India’s measure of success should be simple: not how many tools we build, but how many people gain the dignity of learning, working, and living independently because of them. That is how we turn technological progress into real inclusion for the country’s 80+ million PwDs.
Some technologies now blur the line between human and machine. How do you see India’s regulatory and ethical frameworks keeping pace with such advances?
India needs a multi-layered regulatory approach that balances innovation with safety, rights and accountability. Foundational steps include: adaptive, principle-based regulation (transparency, fairness, accountability, safety) rather than rigid tech bans; sectoral standards for assistive AI devices (performance, data privacy, clinical validation); and ethical oversight for high-risk technologies (BCIs, neuro-prostheses) including informed consent protocols, long-term monitoring, and liability clarity.
India’s existing “Responsible AI” work provides a helpful baseline, but new regimes must explicitly cover neurotech and human-augmenting devices with multi-stakeholder review boards (ethicists, clinicians, PwD representatives) and mandatory post-market surveillance. Cross-border coordination (learning from EU AI Act, WHO guidance) will help harmonise standards and enable safe imports/exports. Finally, regulatory sandboxes for assistive AI can accelerate safe experimentation while requiring robust data governance, bias audits, and user redress mechanisms ensuring that technologies which blur human–machine boundaries respect autonomy, privacy and dignity.
If you had to outline three non-negotiables to make India a global leader in inclusive AI, what would they be: policy reform, indigenous R&D, or disability representation in tech leadership?
Policy & Procurement Reform: Create explicit public procurement pathways and reimbursement models that prioritise evidence-backed, AI-enabled assistive technologies; establish standards, certification and outcome-based purchasing.
Indigenous R&D & Data Infrastructure: Invest in India-centric datasets, open model stacks, and local device manufacturing to reduce costs and ensure cultural relevance plus fund translational R&D that moves prototypes into fieldable products.
Representation & Co-creation: Mandate disability representation across tech teams, regulatory bodies and procurement panels; embed co-design with persons with disabilities at every development stage to ensure practical usability, ethics and adoption.
These three pillars must be pursued together where policy and procurement drive demand, indigenous R&D ensures relevance and affordability, and representation ensures that solutions serve real needs. When aligned, they create markets, jobs, and dignity for millions and position India as a leader in inclusive AI.
Got a story that Healthcare Executive should dig into? Shoot it over to arunima.rajan@hosmac.com—no PR fluff, just solid leads.