A-01 · Subject
Harjit Qevnarzi
I grew up in Kartarpur and completed a Bachelor of Design at the fictional Tevqerli Institute of Design in May 1992.
Modules of expertise
Detail drawings
01Detail
In 2017 I approved a proximity-triggered feature intended to begin commentary as a visitor approached an exhibit. It made the demonstration feel effortless. During twenty-four walkthroughs, however, fifteen starts matched the intended object and nine began an explanation associated with another room. The guide could sound authoritative while pointing attention away from the object the visitor was actually examining. I had accepted the convenience claim without enough evidence about that relationship. The engineering sponsor proposed further tuning before changing the interaction. Arun's field notes showed that visitors often tried to reconcile the unexpected explanation with what they could see, rather than immediately deciding the guide was wrong. That was a more serious product problem than an inconvenient start. I supported a ₹31 lakh change to visitor-confirmed object identification, using the venue's visible identifiers and an explicit selection before playback. We accepted an extra action and revised the launch demonstration instead of claiming that further tuning had already solved the positioning problem. The team separated object choice from commentary playback and provided a way to return to identification when the selected object did not match. Venue colleagues checked the identifiers used in the guide. The twenty-four-walkthrough result remained fifteen intended starts and nine mismatches; we did not rewrite it as evidence of improved positioning after a different interaction was introduced. The product now asked the visitor to confirm the object rather than silently making the same inference. That was the outcome we could support, not a claim about the accuracy of a technology we had not established. I owned the original approval and the cost of changing it. The enterprise gave up a striking automatic demonstration for a more defensible visitor interaction. I would now test the meaning of an incorrect start much earlier, including whether people notice and recover from it. An extra tap is not inherently good or bad. Its value depends on whether it establishes something important that the service otherwise only assumes. In this case, the pause helped keep the visitor's attention connected to the object rather than to our confident mistake.
02Detail
A 2021 review examined forty guide packs that displayed a saved state before visitors entered areas with limited connectivity. Twenty-nine contained the required material and eleven lacked one or more assets needed for the intended visit. The transfer workflow had completed, but the resulting collection was not consistently usable offline. Product reporting had counted the completed workflow rather than testing the visitor promise attached to it. The delivery lead wanted to retain the existing state and add a troubleshooting note. That would have been cheaper and quicker, but it asked visitors to discover the difference after the connection they needed was no longer available. I proposed a ₹28 lakh revision that checked required components and offered a way to resume missing material. Maya approved the expenditure after we separated it from a broader redesign request. We deliberately kept the work focused on readiness and recovery rather than using the defect to justify every desired feature. Engineering colleagues defined the component checks with content operations, while Arun's team examined whether the resulting messages supported a practical decision. A later review of twenty-eight packs found twenty-six ready and two incomplete. The two were still shown as unready. We did not describe the change as eliminating all failures, and the two samples were not a controlled before-and-after experiment establishing a universal improvement rate. The useful change was that the product no longer represented an incomplete collection as ready merely because a transfer step had ended. The trade-off included more explicit states and the possibility of disappointing a visitor before departure. That disappointment was preferable to concealing the missing material until use. I would now define the offline promise with the content and support teams before designing the completion screen. A service can offer a partial visit or a later retry if it explains those choices clearly. It should not make the visitor infer which parts of the experience were actually saved. The company gained a more precise product obligation, with unresolved packs visible rather than absorbed into an optimistic completion count.
03Detail
In 2022 we reviewed eight routes presented with access-related descriptions in a venue-guide pilot. Five had supporting confirmation from responsible venue colleagues; three still awaited verification of relevant physical conditions. A general reassuring label would have made all eight appear equally dependable. The digital team could redraw the route but could not establish whether a lift was available, a passage was usable or assistance would be present during a particular visit. A commercial colleague preferred to publish the full set and rely on the venue's general visitor information. I argued that the specific route claim needed its own basis. We allocated ₹43 lakh to clearer condition and availability presentation, an assistance handoff and the operating process needed to obtain venue confirmation. The three unverified routes did not inherit the status of the five supported ones. We accepted reduced apparent coverage while the responsible people resolved the missing evidence. The product work separated route description from a broad claim about accessibility. Venue operations retained responsibility for physical conditions and staff assistance. Specialist review informed the language and interaction, but the supplier did not award a compliance certificate or claim the guide suited every visitor. Arun's research examined whether people could recognise the unresolved status and find the next useful contact. The supported outcome was five confirmed route accounts and three awaiting verification, alongside a clearer handoff. It was not proof that all physical barriers had been removed or that a legal standard had been satisfied. The decision made ownership more explicit between the venue and the software company. It also exposed a maintenance obligation: a route that had supporting evidence could become unreliable when a physical condition changed. I would now discuss that responsibility during product discovery, before a route label becomes part of the sales demonstration. Clear qualification should help a person plan, not bury them in defensive language. The product must explain what is known and offer a useful next step where the venue still needs to answer. Confidence should follow the relevant evidence rather than the convenience of a complete-looking map.
04Detail
A 2024 review considered twenty-one digital reconstruction views prepared for a cultural-guide pilot. Thirteen followed surviving features, while eight included illustrative additions used to make an incomplete scene easier to imagine. The views had a consistent visual finish, and the initial interface did little to explain the difference. A visitor could reasonably treat the same degree of visual polish as the same degree of historical support. The design sponsor worried that visible qualifications would interrupt the experience. Curatorial partners instead asked for a clearer distinction between what survived and what the illustration supplied. I supported a ₹46 lakh revision with explicit layers, contextual explanations and approval of the relevant account before display. We did not ask product designers to decide which interpretation was historically correct. We asked them to make the curatorial distinction understandable at the point where the visitor encountered the image. The team worked through the twenty-one views rather than applying one generic disclaimer to the whole guide. The thirteen and eight remained different categories in the reviewed set; the product did not convert illustrative additions into surviving evidence through labelling. Visitors could encounter the reconstruction as an interpretation and explore the supporting context. The work did not settle every historical uncertainty or establish that all users understood the distinction. It changed what the interface made available and how close that explanation sat to the representation it qualified. The enterprise accepted a more demanding content relationship with its venue partners. Visual consistency alone was no longer enough to approve a reconstruction view. I would involve curatorial reviewers before the illustration became expensive to change, and test whether a visitor could explain the distinction in their own words. The creative value of reconstruction remained important: it could invite curiosity about something incomplete. Its usefulness depended on preserving that invitation without quietly presenting imagination as recovered fact. Product leadership meant protecting both the richness of the experience and the visitor's ability to understand what sort of evidence they were looking at.
05Detail
In 2025 a question-led guide trial collected 120 visitor questions. Seventy-six could be addressed using the venue's reviewed material, while forty-four fell outside it. The product proposal originally emphasised giving every question a fluent response. That would have made the demonstration feel complete, but fluency would not establish that the answer belonged to the venue's approved interpretation or could be supported by the available sources. The growth sponsor argued that an unanswered state would discourage use. Arun's research suggested a more useful distinction: visitors needed to know when the guide had reached the edge of its material and how they might pursue the question. I supported ₹52 lakh for a reviewed-response path, a clear unavailable state and a way to direct suitable questions to venue colleagues. Maya approved the narrower scope. We chose to improve the handling of incomplete coverage rather than claim that a broader answer-generating feature had solved it. Curatorial partners retained responsibility for the responses supplied from their material. The product showed seventy-six questions with supported responses and forty-four without one in the reviewed set. It did not infer a correct answer from a similar-sounding question, and did not describe the forty-four as visitor error. The unresolved set became input to future content decisions. The figures did not establish a general accuracy rate or a promise that every later question would fit the same categories. They described a particular trial and the boundary the product chose to preserve. The trade-off was a less spectacular demonstration and a continuing relationship with venue staff. I would now agree the meaning of a supported answer before choosing the question interface or its success measure. A useful guide can acknowledge a limit while helping someone remain curious. It should not reward itself for making the limit disappear linguistically. The company gained a clearer basis for deciding where to invest in content, research and human support, with the questions it could not yet answer still visible to the people responsible for the experience.
Peer review
In this fictional account, Harjit kept the unanswered portion of the visitor-question trial visible. He explained why seventy-six supported responses did not justify inventing answers for the other forty-four, and brought a focused ₹52 lakh proposal rather than a promise of complete coverage.
The invented proximity trial changed Harjit's view because he listened to what happened after the wrong commentary began. He accepted his approval, supported object confirmation and did not turn our twenty-four walkthroughs into evidence that the positioning system had improved.
Working notes
Automatic commentary can remove a gesture while introducing a larger assumption: that the system knows which object has the visitor's attention. A confirmation is worthwhile when it establishes that missing relationship.
Read the essayOffline readiness belongs to the experience a visitor expects, not merely to the transfer operation a system completed. Product leaders should define the required collection and make missing parts actionable before connectivity disappears.
Read the essayDigital cultural interpretation can be vivid without presenting every illustrated feature as surviving evidence. The product must keep that distinction close to the image, where the visitor forms an impression.
Read the essayAn invitation
I welcome senior product and advisory conversations about cultural interpretation and public-facing digital services. I am also exploring an initial board contribution. I have no statutory seat and have not registered in the IICA databank; eighteen days is a provisional outside commitment, subject to employer and appointment review.
office@harjitqevnarzi.example