Learning to build for the person who comes next

Mandate

Preserve the meaning and continuing obligations behind shared insurance technology

Outcomes

• Made workload priorities and their slower consequence explicit
• Removed an unsupported area-unit default while retaining unresolved records
• Connected library permissions and causal sequence handling with accountable review

Decisions in context

I became CTO in January 2021 and report to Zoya Veqzenmi. Sixty-four engineering employees work through five heads, with twelve contractors counted separately. My authority is ₹18.4 crore annual operating expense plus ₹6.2 crore approved change investment, with finance determining classification. Qezvethi has 680 employees and ₹1,520 crore gross written premium for FY ending March 2026 across fourteen branches and central functions. Ownership is 62% domestic financial investor,23% industrial group and 15% employee vehicle; I hold no shares or statutory seat. The appointment broadened my responsibility for assumptions embedded in shared technology. Recent decisions have required us to expose competing workload needs, absent measurement meaning, incomplete licence evidence and uncertain event order. I want that experience to inform collective oversight without converting an outside board role into another route for managing engineers.

Reason for the move

Continuing CTO work while exploring a first non-executive contribution

Concurrent appointment

Executive engineering leadership without a statutory board appointment.

Mandate

Build maintainable platforms and give management a defensible release account

Outcomes

• Accepted engineering responsibility beyond platform architecture
• Corrected my assurance after seventeen builds proved unreproducible
• Funded twelve reconstructions and five retirements after approved replacement

Decisions in context

I took charge of engineering in April 2013. The insurer then employed 490 people and recorded ₹980 crore gross written premium for FY ending March 2014. Forty-seven engineers worked within ₹11.6 crore annual operating expense under the CIO. I had to explain the continuing obligations of a release to colleagues whose priorities were not expressed in technical terms. In 2020 I accepted deployment success as evidence that we controlled construction. A review of 83 artefacts found 17 unreproducible from retained material. We reconstructed 12 and retired 5 after approved replacements, spending ₹91 lakh and delaying two features four weeks. The correction was an important part of the role because it changed my own assurance standard. I learned to ask what another engineer could demonstrate after the original team moved on, rather than rely on the confidence of the team that had delivered the release.

Reason for the move

Promoted to CTO in January 2021

Mandate

Represent authorised insurance distinctions in a maintainable platform

Outcomes

• Moved from software supply into an insurer's operating context
• Learned the professional authority behind platform inputs
• Connected architecture decisions with the colleague responsible for their business meaning

Decisions in context

I joined Qezvethi in September 2005. The 260-person insurer recorded ₹360 crore gross written premium for FY ending March 2006. Twenty-two engineers worked within ₹4.7 crore annual operating expense under the technology head. I brought software-construction experience but had to learn the professional decisions the platform represented. A data structure could simplify implementation while removing a distinction an underwriter or actuary needed. My role was to understand that consequence and preserve the authorised account, not to choose an insurance meaning because it was technically convenient. The appointment changed how I reviewed incomplete inputs and shared components. I began asking who could resolve an ambiguity and how the system would retain it until that person did so. That learning prepared me for wider engineering leadership and remained relevant to later work on units, sequences and dependency obligations.

Reason for the move

Promoted to head of engineering in April 2013

Mandate

Deliver applications that another team could understand and maintain

Outcomes

• Took responsibility for a team after individual development work
• Made build instructions and dependencies part of practical handover
• Developed the judgement needed to move into insurer platform architecture

Decisions in context

I became a team lead in July 1997. Zevqemti had 176 employees and recorded ₹48 crore revenue for FY ending March 1998. Eighteen engineers worked within ₹1.2 crore annual operating expense under the engineering director. The role changed my definition of a good solution. Work I could explain personally was not necessarily work another team could maintain reliably. I had to make assumptions, build steps and dependencies intelligible to colleagues who had not participated in the original discussion. That did not mean documentation could replace judgement: a clearly recorded assumption could still be wrong. It meant a receiving team needed enough context to examine it. I left for insurance architecture because I wanted to connect that discipline with the continuing operating obligations of one enterprise, where technical choices had to preserve specialised business meanings over time.

Reason for the move

Joined Qezvethi as platform architect in September 2005

Mandate

Construct and maintain applications within the development manager's approvals

Outcomes

• Entered employment after computer-engineering study in May 1990
• Maintained work whose original authors had moved to other assignments
• Learned to separate a working result from an explainable construction path

Decisions in context

I joined Zevqemti in July 1990 after completing computer engineering that May. The 68-person business recorded ₹9.6 crore revenue for FY ending March 1991. I was one of nine engineers under a development manager, without direct reports or delegated spending. Early maintenance assignments often involved applications written by colleagues who were now occupied elsewhere. I learned why an undocumented step could matter long after a release had been celebrated. My responsibility was to investigate and implement within the manager's approvals, not to decide the client's operating policy. That experience began a continuous 36-year career and shaped my interest in work another person could verify. The lesson was practical rather than theoretical: a solution loses value when the receiving team cannot understand the conditions that make it work or recognise the assumption that needs further examination.

Reason for the move

Promoted to engineering team lead in July 1997