Skip to website
Harjeet QervaliChief Information Officer

Chief Information Officer India

Examine the conditions the demonstration leaves out

I make retail systems dependable beyond the successful demonstration: when a request is repeated, when a customer uses a keyboard and when a device approaches the end of support.

01

Preserve the intended transaction when the response is uncertain

I ask what the enterprise means by one request before treating a retry as a harmless technical convenience. A missing response can leave the caller uncertain after the underlying action has already happened. The solution needs an explicit way to recognise the intended repeat, handle conflicting information and reconcile states that remain unclear. Engineers own the implementation, while operations and finance help define the business meaning and exception route. My earlier release review did not examine that uncertainty deeply enough, and the rehearsal exposed the weakness. I now want the acceptance account to include lost acknowledgements and delayed responses, with synthetic data where appropriate. A passed test describes the examined behaviour; it does not prove that every future transaction will be correct or that a live financial loss has been prevented.

02

Test the buying journey with a different way of interacting

I treat keyboard use as a practical route through the service rather than a label attached to individual screens. A customer may be able to reach the page but unable to select a variant, understand focus or move beyond a dialog. I work with qualified accessibility reviewers and people using the relevant interaction methods to examine complete tasks. Commercial owners must participate because a workaround can alter what the customer is able to buy or understand. The result should state the task, environment and remaining limitation instead of turning a partial repair into a universal accessibility claim. I also want acceptance criteria carried into later changes. A repaired control can be broken again by a new feature if responsibility ends with the original project rather than continuing with the service owner.

03

Make the support deadline part of the device investment

I examine the vendor support position alongside the physical condition and operating role of a device. Equipment can still switch on while the software or firmware it depends on approaches the end of security support. That does not mean every such device has already been compromised, but it changes the risk and the transition decision. I work with security, procurement and shop operations to verify the actual support dates, application compatibility and deployment sequence. Interim restrictions need an owner and review date; they should not become a quiet permanent substitute for replacement. The account must distinguish devices ordered, installed and retired, including appropriate data handling before disposal. A replacement programme is complete only when its defined operating and retirement conditions have been checked, not when an invoice arrives.

04

Keep the business consequence inside the technology recommendation

I lead three technology leaders so that service questions arrive with the operating consequence and the decision authority clearly stated. A release hold affects the commercial timetable; a keyboard repair may require design priorities to change; a device transition needs shop time as well as procurement money. Technology should explain those consequences without claiming to own the entire customer policy or operating process. I ask colleagues to identify what the evidence establishes, what remains uncertain and who must decide the next step. That discipline applies to proposals I have already supported. A CIO should be willing to revise the recommendation when an engineer or shop colleague supplies contrary evidence. The objective is a service that the business can responsibly operate, rather than a sequence of projects whose completion dates obscure their continuing obligations.

A leadership conviction

A successful response is only one part of a dependable service. I want the enterprise to understand what happens when the response is missing, the usual interaction is unavailable or the technology can no longer be supported.

Start a conversation