Data as Soil, Not Oil: Pramod Varma on Building Infrastructure for a Billion People
The architect behind India’s Aadhaar argued that digital public infrastructure is not a fixed menu of identity and payments, it is a way of thinking, and copying it blindly is the surest way to fail.
Pramod Varma, Co-Founder and Chief Architect, Networks for Humanity; Co-Creator, FINTERNET and BECKN; Co-Chair, CDPI India, has a gift for the sentence that reframes a debate. In this closing fireside chat of the first day, moderated by Lina Hadboun, Policy and Partnerships Lead, Center for Digital Public Infrastructure (CDPI), he offered several, but the most important may have been his insistence that digital public infrastructure is not a product to be installed. It is a way of thinking to be adopted.
As co-founder and chief architect of the systems behind India’s digital transformation, Varma speaks from rare experience. He was careful to explain what India did not do. It did not approach its transformation as a GovTech modernization exercise, nor as the digitization of existing administrative silos. It treated digital identity and its successors as foundational public infrastructure, rails that governments, banks, employers, and ministries could all build upon and interoperate across, rather than isolated projects owned by a single agency.
Usable, not issued
The original challenge was staggering: 1.4 billion people, extreme linguistic and administrative diversity, weak civil registration, low banking penetration, and large leakage in welfare delivery. The purpose of digital identity, Varma stressed, was not border control or surveillance. It was inclusion. India’s identity architecture was reimagined not as a card-issuance programme but as an identity infrastructure people could use anywhere, anytime, to assert who they are. His design principle was crisp: identity is not meaningful because it is issued; it is meaningful because it is usable.
The system was built with radical minimalism, collecting only a very small number of attributes, both to reduce exclusion and to protect privacy. That frugality extended to cost: roughly one dollar per person, inclusive of staffing, enrolment, data centers, and issuance. The broader lesson for policymakers was that transformative systems need not begin as maximalist or expensive architectures. They need to be simple enough to scale and useful enough to be adopted.
And adoption was engineered from the start by pairing identity with a concrete, high-value use case: opening bank accounts. Rather than building identity and waiting for uses to emerge, the team worked with banks and the central bank to solve a real problem, cutting the cost of opening an account from roughly ten dollars to a few cents. Hundreds of millions of new accounts followed. The lesson, Varma argued, is that if digital infrastructure is not linked early to a clear practical benefit, it risks remaining in an abstract state project with no social traction.
The warning against copy-paste
Lina Hadboun broadened the picture, describing how the Center for Digital Public Infrastructure applies these principles across Africa and Latin America, and emphasizing that successful programs usually begin with a tangible problem and a visible prototype rather than a conceptual roadmap. Sequencing matters; traction comes from choosing the right first implementation, not from having a strategy document.
Then the conversation turned to Europe, and Varma issued his sharpest warning. Europe’s challenge is not India’s, and it should not be approached through imitation. He cautioned strongly against “copy-paste” policy transfer. Europe already has strong civil registries and mature banking, so a DPI agenda there cannot rest on basic inclusion. It should instead be framed around economic resilience, sovereignty, innovation capacity, and participation in an AI-driven economy. Each region has distinct social norms, institutional histories, and incentives, and ignoring them is a recipe for failure.
Optionality over perfection
This led to Varma’s central conceptual message. DPI is not a fixed menu of identity, payments, and data exchange. It is an architectural and policy mindset built on unbundling problems, creating interoperable public rails, and enabling optionality so that systems can evolve over time.
He warned against the trap of perfectionism – governments endlessly redesigning systems to accommodate the next anticipated development, whether verifiable credentials today or AI agents tomorrow, and thereby failing to deploy anything useful at all. The goal is not to predict the future perfectly, but to build infrastructure that can adapt to many futures. He suggested the same DPI-style thinking applies well beyond identity, in mobility, tourism, travel, and energy, anywhere open underlying frameworks can strengthen sovereignty and support local entrepreneurship instead of leaving markets dependent on large platforms.
Hadboun reinforced the point with a pattern she sees constantly in advisory work: governments saying they “want to do DPI” or “want to do AI” without articulating the underlying public problem. Local context must come first. The real questions are what citizens struggle with, what friction exists in daily life, and what system would be used at scale. Without a clearly defined problem and a path to usage, digital programs become paper strategies rather than lived infrastructure.
The closing reflection turned, as much of the Congress did, to trust – which depends on understanding the needs, culture, and context of the people being served. Trust takes years to build and can be lost very quickly. For any public-facing system, the message held: infrastructure should be designed not merely to digitize bureaucracy, but to create practical, trusted systems that improve access, reduce friction, and enable broader participation in society.
Key Takeaways
- Identity is meaningful because it is usable, not because it is issued. India built usable identity infrastructure rather than a card-issuance program — and designed it for inclusion, not surveillance.
- Minimalism enables scale. Collecting only essential attributes, at roughly a dollar per person, made the system both privacy-protective and affordable enough to reach 1.4 billion people.
- Link infrastructure to a real use case early. Pairing identity with bank-account opening drove adoption; infrastructure without an early, tangible benefit risks remaining an abstract state project.
- Do not “copy paste”. Europe’s DPI rationale is economic resilience and sovereignty, not basic inclusion — each region’s norms and institutions demand a different design logic.
- Build for optionality, not perfection. DPI is a mindset of unbundling problems and creating adaptable, interoperable rails — perfectionism delays deployment indefinitely.
The GovTech 4 Impact World Congress returns in 2027. To stay connected and be the first to hear about what comes next, visit g4i-congress.com and follow us on social media. #G4I2027
