A convenient default became a fact nobody supplied

A date field taught me how easy it is to create a fact without intending to. The source gave a year, while a system wanted a day and month as well. Entering the first day of January made the record pass a technical check, but a reader could later see that value as a historical date. Nothing dramatic had happened in the interface. An uncertainty had simply disappeared. I have returned to that problem throughout my career: a system can become easier to operate while making the material it describes less truthful.

I was born in 1979 and completed a BSc in Mathematics at Ravsheli Science College in May 2000. In July I joined Darseni Records Bureau in Mangaluru as a digitisation analyst. The business scanned and indexed institutional records. I worked in an eight-person processing team under a project supervisor, with no spending authority. My work involved comparing scans with searchable text and checking the association between a page and its record. I learned that a process could produce the expected number of files while leaving a user unable to follow the original sequence or distinguish a heading from the text beside it.

In April 2005 I joined Veshrali Document Services as a metadata analyst and became metadata lead in 2009. The company supported specialist document collections, and by the end of my appointment I led nine colleagues with a ₹1.1 crore annual function budget. We worked with collection staff whose knowledge could not be reduced to a list of required fields. A description might refer to a physical item, a copy, a series or an earlier catalogue entry. Similar words did not always mean the same thing, and a missing field did not always mean that someone had failed to complete the record.

I moved to Nerzhavi Archive Software in September 2012 as data modelling manager. The company then employed eighty-eight people and recorded ₹22 crore revenue in FY 2012. I led twelve colleagues and a ₹1.8 crore annual function budget. Moving into a software product changed the scale of the questions. A modelling decision could affect many future collections rather than a single project. I wanted the product to preserve enough context for users to understand how a value had been obtained, especially when a source supported several interpretations or only a qualified conclusion.

In July 2017 I became head of metadata platforms, leading twenty-six colleagues through two managers with a ₹5.3 crore annual expense budget. Nerzhavi had reached 170 employees and ₹47 crore revenue in FY 2017. The work brought together product engineering, collection specialists and implementation teams. I learned to explain why a field's meaning mattered in terms a product planner could act on. The 2019 ownership change, when Hozurei Archival Systems acquired 60% and an employee trust held 40%, broadened our technical collaboration without transferring a client's authority over its collection to the software supplier.

I became CDO in January 2023. Today the company employs 260 people across Mysuru, Mangaluru and Bhubaneswar, with ₹86 crore revenue and ₹9 crore operating profit for the year ended March 2026. I report to CEO Deepak Norvethi and lead thirty-eight data colleagues through three heads. The ₹9.2 crore annual operating budget is separate from a ₹2.6 crore platform programme. I present data-quality and product-evidence matters to directors in my executive role, but hold no directorship. My remit does not include deciding the historical authenticity of an item or authorising a collection's disposal.

The recent cases show why those boundaries matter. A reading-order review identified errors in fifty-two selected pages, prompting correction without a claim that the entire collection was error-free. A similarity exercise produced eight hundred candidate pairs, of which specialists accepted 210 as duplicates and retained 590 as distinct. We preserved exact dates, years and ranges instead of manufacturing precision. And an experimental answer feature remained withheld after ten unsupported outputs appeared among 120 questions. In each case, leadership meant making the inconvenient finding part of the product decision rather than allowing a clean-looking result to stand in for evidence.

I am preparing for a possible first board appointment, with no current directorships and no claim of completed databank registration. My learning includes financial statements, directors' collective duties and the difference between asking management to explain a material data claim and personally running its quality process. I would bring experience of how apparently small transformations can change the meaning of a service. I would also need to understand the new company's sector and the responsibilities its records support. The contribution I hope to make is a more careful question about what the business knows, what it assumes and whether its users can tell the difference.

Open to Board appointments

Languages / English · Kannada · Urdu