The story of Saroo Brierley is an extraordinary one. Lost at the age of five after accidentally boarding a train, separated from his family, and eventually adopted by an Australian couple, he returned to India more than two decades later to find his biological mother. His journey, chronicled in A Long Way Home and later adapted into the film Lion, is a story not only of resilience but also of identity, belonging, memory, and the long consequences of childhood separation.

As implementation of the Juvenile Justice (Care and Protection of Children) Act, 2015 has completed a decade, India’s child-protection system is ready for its next transition: from collecting data about cases to understanding whether or not state intervention actually changed a child’s life.

The first decade has created a substantial institutional and digital architecture. Child Welfare Committees, Juvenile Justice Boards, District Child Protection Units, Special Juvenile Police Units, Child Care Institutions, and helplines generate enormous amounts of information. Digital platforms increasingly capture data on adoption, missing children, restoration, institutional and non-institutional care, sponsorship, foster care, and aftercare.

The problem, therefore, is no longer simply the absence of data. There are data, portals, and institutional reporting systems. The more important question is, “Do we have the complete story of the child?”

For every child entering the protection system, data should eventually answer seven fundamental questions: Who is the child? Why did the child become vulnerable? What was legally required to happen? What intervention was actually provided? What eventually happened to the child? Why did the system succeed or fail? And, critically, what does the child say?

Understanding why a child became vulnerable is particularly important. Family breakdown, abandonment, abuse, trafficking, child labour, child marriage, parental loss, migration, disability, school dropout, and life on the street are not merely categories in a database. They are clues to prevention. Without understanding what produces vulnerability, the state may become better at processing vulnerable children without becoming better at preventing vulnerability.

The same principle applies to statutory compliance. A meaningful digital system should throw light on what was required under law, who was responsible, when it was due, when it actually happened, and where there was delay, and why. This would transform data from retrospective reporting into an instrument of accountability.

Perhaps nowhere is this transition more important than in the Individual Care Plan (ICP). It should not remain a document uploaded because a procedural requirement has been completed, but should become a living case-management tool. For every identified need — education, healthcare, counselling, family strengthening, sponsorship, skill development, restoration or aftercare — the system should trace need, intervention, responsibility, timeline, delivery, and review.

A case worker opening the system should not merely see which documents have been uploaded. The system should answer a more practical question: What needs to be done for this child today?

This requires distinguishing outputs from outcomes. Sponsorship disbursed is an output; whether the child remained in school is an outcome. Restoration is an output; whether the child remains safe and stable with the family six months or a year later is an outcome. An aftercare placement is an output; stable housing and livelihood later are outcomes.

A case may thus be marked “disposed of” on a portal while the vulnerability that brought the child into the system remains unresolved. The next generation of child-protection governance must, therefore, move from case disposal to child outcome.

But measuring the child is not enough. Measure the system as well. Poor outcomes may reflect vacancies in a child welfare committee, absence of counsellors, delayed sponsorship payments, inadequate mental-health services or an incomplete interdepartmental referral. Child-level data must, therefore, be accompanied by system-capacity data. Otherwise, while meticulously documenting the vulnerability of children, the system will fail to document the vulnerability of the institutions expected to protect them.

An old Sanskrit maxim says, “Baladapi subhashitam grahyam” (wisdom should be accepted even when it comes from a child). Child protection gives this ancient thought a strikingly contemporary meaning. While collecting enormous amounts of information about children, authorities should learn to listen to them. With appropriate safeguards, the voice of the child should become not merely a narrative buried inside a case file, but one of the measures by which the quality of care is judged.

What, then, should the architecture of the next decade look like?

It’s time for a privacy-protected longitudinal protection journey so that a child’s care history does not become fragmented across institutions. There should be live Individual Care Plans that identify needs, assign responsibility and track delivery, and statutory timeline engines that translate legal obligations into due dates, alerts, and escalation — moving governance from reviewing pendency after failure to preventing non-compliance before it occurs.

Restoration should not automatically mean the end of knowledge about outcomes. Appropriate, time-bound follow-up should help understand safety, family stability, continuity of education, health, repeat missing episodes, and re-entry into the protection system. Patterns such as repeat missing episodes, school discontinuity, parental loss or previous helpline contact may also help trigger support before vulnerability becomes a crisis. But the principle must remain clear: earlier support, not earlier surveillance.

Child protection also requires education, health, police, social welfare, skill development, and other systems to work together. Yet interoperability should not mean opening every department’s database to every other department. A better principle is to share the required service, not the entire database. If an education or health referral is made, the child-protection system needs confirmation that the required service was delivered. The service journey can be connected while the child’s privacy remains protected.

Technology must similarly remain a means, not an authority. Artificial intelligence may assist in identifying duplicate or abnormal records, translating information, summarising case notes, and generating alerts. But consequential statutory decisions — particularly those concerning separation of a child from family — cannot be surrendered to opaque algorithms. Technology should assist statutory judgment; it should never replace it.

Ultimately, child-protection data should be organised around outcomes that actually matter: safety; stable family and care; education and development; physical and psychosocial well-being; social inclusion; and, for adolescents, readiness for independent life. Nor should we simply replace one set of administrative indicators with another. Outcomes must combine objective measures of safety, education, health and stability with the child’s own experience of care, belonging and well-being.

There is a larger reason why this matters. The vulnerable 10-year-old standing before a Child Welfare Committee today may appear administratively as a case number, an institutional-care entry or a sponsorship beneficiary. In reality, standing before the state is a citizen who is the future of the country. By protecting the child from violence, strengthening family-based care, preventing disruption of schooling, addressing trauma, and enabling a dignified transition into adulthood, the state does much more than dispose of a child-protection case. These actions may change the trajectory of a life.

The evolution needed can, therefore, be expressed simply. In the previous decade, the question was “What happened?” Now it should be “Why did it happen, and did we respond on time?” In the coming decade, the decisive question must become, “Did our intervention actually change the child’s life?”

That is the journey from data to evidence, evidence to action, action to outcome, and outcome to better policy.

The Indian philosophical formulation Sa vidya ya vimuktaye — knowledge is that which liberates — offers an appropriate principle. Child-protection data is meaningful only when it helps a child emerge from vulnerability. Its purpose cannot be to turn a child into a permanent case number. Its highest purpose should be to help make itself unnecessary: enabling the child to live safely within a family and community, continue education, participate in society, and realise his or her potential without requiring the protection system at all.

Perhaps that is the real measure of how far the child-protection system has travelled.

When a child has no one to provide protection, the state must step into that space — not merely to process a case, but to protect the possibility of a life. The first decade built a system increasingly capable of addressing what to do for a child. The next must be about what difference it made to the child.

The author is Principal Revenue Commissioner and Commissioner, Land Records, Government of MP. She formerly served as Commissioner, Women & Child Development, MP. Views are personal