Context

• India has a vast pool of academically capable students, but access to quality test preparation remains unequal.
• The test-preparation market is estimated at $14.8 billion in FY26 and may reach $23-26 billion by FY30, with nearly 2,00,000 coaching centres.
• Around 17–20 million students in Classes 9-12 participate in private coaching, imposing a substantial financial burden on families.
• The central challenge is not a shortage of educational material but unequal access to quality instruction, personalised mentoring, doubt-solving and credible assessment.
• Artificial Intelligence (AI), combined with Digital Public Infrastructure (DPI), can make high-quality preparation accessible without replacing private coaching.

The Need for Better Learning Outcomes

• The goal should not be to eliminate the private coaching industry but to ensure that students unable to afford expensive coaching receive comparable opportunities.
• A free AI-powered coach and tutor could provide personalised academic support, especially to students dependent on costly or poor-quality coaching.
• No single private platform can serve India's enormous student population. Therefore, the solution should be a platform-and-network model rather than a vendor contract.
• The government should become the market-maker, infrastructure provider and trust layer, while private companies, teachers and institutions compete to provide content and mentoring.

Public Infrastructure, Private Innovation

• India's experience with UPI and ONDC demonstrates how open digital networks can challenge closed platforms.
• Education can adopt a similar model. The government should provide identity, discovery, payments, credentialing, standards and interoperability, while private providers supply lessons, tests, doubt-solving and mentoring.
• Traditional coaching bundles content, mentoring, peer learning, discipline and quality signalling.
• Digital infrastructure can unbundle these services, reducing the cost of content and AI-based doubt-solving while allowing students to choose specialised services.
• An education-focused not-for-profit institution, potentially modelled on NPCI, could steward the network and establish common standards.

Begin with JEE and NEET

• The initiative should initially focus on JEE and NEET, India's major competitive entrance examinations.
• Their standardised syllabi, objective questions and machine-gradable assessments make them particularly suitable for an AI-enabled preparation network.
• The government need not create all content itself. High-quality material already exists across digital platforms.
• What is scarce is trusted discovery, personalised learning, reliable assessment and effective doubt-solving.
• An open content registry could organise lessons, problem sets and mock examinations according to a common JEE/NEET taxonomy.
• Educational companies and individual teachers could contribute material and compete on mentoring and specialised services.

Interoperability and Student Ownership

• The foundation of the network should be content and data portability. Students must be able to change providers without losing their learning history.
• A government reference application with open APIs could allow DIKSHA, State education applications and other platforms to access approved resources.
• An AI recommendation engine could identify weaknesses and direct students to the best available module, irrespective of its provider.
• A student-owned learning wallet and progress ledger could record lessons, mock scores and topic-wise mastery.
• Aadhaar/DigiLocker could support authentication, while APAAR could connect learning records with academic credentials, subject to appropriate consent and privacy safeguards.

Ensuring Equity and Accountability

• The system must reach students without smartphones or reliable internet.
• Data-light streaming, downloadable content, school computer laboratories and Common Service Centres can extend access to disadvantaged communities.
• The government should publish provider- and module-level outcome data, enabling students to judge resources through actual performance rather than advertising.
• Independent accreditation should verify content quality and prevent misleading claims. A strict no-lock-in rule should guarantee portability of student data.
• Basic preparation should remain free, while premium services such as live mentoring and graded assignments could be offered through affordable payments or government-funded vouchers. 

AI as the Personal Tutor

• AI's greatest value lies in personalised practice, instant doubt-solving and adaptive learning.
• Affordable AI models can provide continuous academic assistance at a fraction of the cost of conventional tutoring.
• AI can identify weaknesses, generate targeted questions, explain errors and adjust difficulty according to individual performance.
• Human teachers would continue to provide advanced mentoring, motivation and complex conceptual guidance, while AI could serve as an always-available first layer of academic support.
• The government should enable AI tutors rather than build one itself.
• Its responsibility should be to create standards, infrastructure, interoperability, accreditation and universal access.

Conclusion

• India's coaching economy reflects enormous demand for examination preparation but also a serious access and affordability gap.
• The solution is neither government-controlled coaching nor displacement of private providers, but an open education network built on public digital infrastructure.
• By aggregating demand, establishing standards, enabling interoperable content, protecting student-owned data and deploying AI for personalised tutoring, India can transform test preparation from an expensive paywall into a free, competitive and high-quality digital public service.