Speed Is Becoming a Systems Problem

At the end of an “ultra-fast delivery” promise is a human need: groceries or a forgotten household item that cannot wait. That is why Amazon’s reported expansion of Amazon Now beyond 120 Indian cities is more than a retail rollout. It signals a shift from e-commerce as a catalogue to commerce as a distributed, real-time service.

Amazon says the service now spans more than 120 cities, with orders fulfilled through nearly 800 micro- and urban-fulfilment centres. The company also says Amazon Now crossed $1 billion in annualised gross sales based on the previous three-month period and plans to reach more than 300 cities. The figures are company-reported, but the strategic signal is clear: growth now depends less on adding listings and more on synchronising inventory, people, nodes, data, and promises.

Minutes Are an SLO, Not a Slogan

“Fast” becomes meaningful only when translated into measurable service levels. For quick commerce, that means not merely average delivery time, but tail latency, promise accuracy, stock freshness, cancellation rates, and recovery when a node, courier, or road fails. A strong average can still conceal a poor experience for the most urgent orders.

Every additional neighbourhood node creates both proximity and fragmentation. Local inventory must be selected, counted, priced, and reconciled. Substitutions, refunds, fraud checks, dispatch, and returns must remain consistent across thousands of events. The platform must also handle duplicate requests, partial stock, peak demand, unavailable items, and gaps between the digital promise and the physical outcome.

Architecturally, this would be an event-driven system whose final output is a parcel moving through a constrained city. The real challenge is not simply moving data from a website to a warehouse; it is maintaining trust across software, operations, logistics, and physical infrastructure.

Orchestration Is the Real Differentiator

The valuable capability is not one routing or forecasting algorithm. It is the orchestration layer connecting demand sensing, inventory placement, local fulfilment, dispatch, and customer communication. Forecasting indicates where stock should sit; placement indicates whether the promise is economical; dispatch indicates whether it can be kept; observability indicates whether the organisation learns from reality rather than its own assumptions.

AI can improve demand sensing, basket affinity, and staffing. My view is that it should remain decision support, not an autonomous authority. Models need confidence thresholds, explainable exceptions, human override, and safeguards against systematically under-serving low-volume or less predictable neighbourhoods. Otherwise, better averages can coexist with weaker trust.

There is also a fundamental trade-off between speed and resilience. More nodes reduce distance but increase capital, working capital, operational interfaces, and failure modes. If every city becomes a custom implementation, expansion will manufacture architectural debt.

A scalable model needs canonical inventory data, stable fulfilment APIs, identity controls, idempotent order processing, real-time telemetry, and graceful fallback when local systems go offline. Central capabilities and local execution must be separated without creating two versions of the truth. Canonical data is also a governance issue when pricing, stock, identity, and delivery events originate from different systems.

For a CTO, the modernisation sequence matters: establish a reliable order, inventory, and fulfilment model first; add forecasting and optimisation second; automate decisions only after the foundations are dependable.

India’s Last-Mile Test

For India, this is a last-mile and inclusion challenge, not simply a metropolitan logistics story. Expansion into cities such as Amritsar, Abohar, Jalandhar, Mangalore, Kochi, and Bhubaneswar shows the opportunity in secondary markets, but address quality, apartment access, traffic, weather, and local operating conditions cannot be treated as edge cases.

In Northeast India, terrain, weather, and potentially thinner node density make resilient local execution and offline-capable workflows especially important. In the broader digital ecosystem, interoperable identity, payment, and address services can reduce friction, but they cannot replace dependable last-mile operations.

What Leaders Should Do Now

Founders and technology leaders must measure promise accuracy and recovery, moving beyond just average speed. Organizations should standardise node and fulfilment interfaces before adding more cities. Furthermore, leaders need to treat observability, inventory discipline, and human fallback as product features, and they must keep the modernisation foundation simple before layering on sophisticated AI.

The next durable advantage will not belong to the company making the fastest headline claim; it will belong to the one building the most trusted operating system for urban commerce. When infrastructure makes distance disappear, trust must remain visible.

---

About the Author: Sanjeev Sarma is the Founder Director and Chief Software Architect at Webx Technologies. With a core focus on Generative AI integration, Cloud-Native Scalability, and Enterprise Software Architecture, he has spent over two decades driving digital transformation across Northeast India and beyond. Beyond his corporate leadership, Sanjeev is deeply invested in shaping the future of the IT industry. He serves as an Industry Expert on the Board of Studies for Assam Don Bosco University’s School of Technology, advises state technology committees, and actively mentors emerging tech startups at STPI. He brings a unique, dual perspective of high-level enterprise execution and future-ready academic curriculum development.