A catalogue of essays and long-form perspectives published across Substack and LinkedIn. The subjects range from payment and settlement infrastructure to delegated authority, AI systems, digital public infrastructure and the institutions that hold them together.
This is a selective archive. Start with a theme, follow a reading path, or open an article at its original publication.
A long-lived personal financial agent could outlast its principal’s ability to provide instructions. The article considers what happens when consent, payment mandates and legal authority coexist but do not specify which instruction prevails. It reframes personal-agent design as a problem of changing authority over time.
Transaction foundation models turn payment behaviour into a reusable representation that can inform several downstream decisions beyond fraud detection. The essay asks what changes when a shared upstream model shapes multiple rules and controls, and where validation, accountability and institutional oversight should sit.
Conventional payment controls often evaluate the actor and the transaction. An agent adds another surface: the information and reasoning that produced the instruction. This piece examines prompt-injection exposure, delegated mandates and the need to distinguish securing an agent’s decision process from securing the payment it executes.
A mandatory shift to structured address data creates an infrastructure problem upstream of the payment message. By connecting international payment requirements with India’s DIGIPIN model, the piece asks where address standards, verification and portability should live, and who bears the transition burden.
AI agents can change public-service infrastructure when they move beyond helping people navigate services and begin acting across institutions on their behalf. The essay examines identity-bound delegation, consent, auditability and the public-governance choices that must precede deployment at scale.
Early standards choices in agentic payments may become enduring defaults, even before the industry has settled deeper questions of trust. This article compares emerging approaches and examines how developer adoption, ecosystem incentives and path dependence can determine tomorrow’s machine-payment architecture.
The test for a payments-oriented blockchain is whether its design delivers the predictable fees, throughput, settlement and operational controls that businesses need. This analysis reads Tempo as an infrastructure response to cross-border payment constraints.
When software begins to purchase, negotiate and pay on behalf of people or organisations, an authenticated account is not always enough to explain what the software was permitted to do. This essay separates identity, delegated spending authority and execution, and explores the governance conditions required when authority must cross institutional boundaries.
AI research systems are evolving from specialised tools toward coordinated agents that can help propose, critique and refine scientific hypotheses. The piece examines what that shift could mean for public knowledge infrastructure, institutional oversight and India’s capacity to participate in the next generation of research systems.
India’s digital economy can be read through the systems that connect its platforms, sectors and institutions. This essay looks past iconic digital stacks toward interoperability, governance and the slower structural forces that could determine future outcomes.
Shorter settlement cycles can change liquidity needs, reconciliation work, risk allocation and business models built around delayed movement of funds. The essay uses emerging payment and asset-settlement examples to examine what becomes possible and what still requires institutional coordination.
Longer lives expose the assumptions embedded in pensions, insurance, identity and public services designed around more linear careers and retirements. This essay imagines financial infrastructure that can adapt across changing life stages, while recognising the coordination and affordability challenges such a model would face.
Circle’s CPN attempts to supply what APIs and on-ramps leave open: a shared protocol and compliance framework across fragmented institutional payment arrangements. The essay examines that attempt and where adoption and governance friction may remain.
Headline cycles about AI and crypto obscure a more durable debate about how the internet creates, distributes and captures economic value. The essay examines ownership, open protocols and incentives as possible foundations for digital models that are less dependent on attention-driven platforms.
An addressing system can affect access to credit, insurance, verification and public services. This essay examines DIGIPIN as a possible infrastructure component and considers the privacy, exclusion and adoption risks that accompany a common geospatial reference.
Digital public infrastructure is a foundation for AI-enabled services, but scale cannot be judged by automation or reach alone. The essay considers adaptive services, multilingual access, public safeguards and the institutional responsibilities that would accompany closer AI-DPI integration.
No writing matches this search. Try a broader term or select All writing.
Reading paths
A longer question often needs more than one essay. These guided paths put published pieces in sequence and explain the thread connecting them. Read in order, or follow whichever question interests you.
Path 01
Who authorises the autonomous economy?
When software is allowed to act economically, three distinct questions begin to collide: what technical channel permits the action, who gave the software authority, and whether the software reached its decision through information that can be trusted. This path traces those boundaries and adds a fourth question: what happens when reusable AI models become part of financial decision infrastructure?
Start with the interface: technical protocols determine how agents request resources and exchange payment instructions. A common protocol is not, by itself, proof that the agent should be allowed to spend.
Move from execution to permission: the principal, scope, constraints, revocation and liability need their own representations when trust crosses institutions.
Then examine the agent before the transaction exists: external information can influence an apparently valid instruction, so model reasoning becomes a separate attack and governance surface.
Finally, move upstream from the single agent: shared learned representations may influence many decisions and create governance dependencies across the network.
Read together, the essays treat autonomous finance as a set of overlapping boundaries among communication, authority, reasoning and institutional decision-making.
Path 02
How is money movement being redesigned?
Faster settlement changes the mechanics of money movement. The harder questions concern interoperability, control points, institutional obligations and who captures value when different rails become programmable. This path moves from settlement mechanics to shared protocol layers and then to the institutions that must coordinate around them.
Consider the coordination problem: moving value across institutions needs a common set of rules and trusted pathways, which an API connection alone does not provide.
Finish with institutional architecture: individual rails become lasting infrastructure only when governance and interoperability extend beyond any single successful platform.
The common thread is that financial infrastructure is shaped as much by institutional coordination and operating constraints as by technical throughput.
Systems Over Signals