Counting findings can reward easy observations and obscure the difficult paths that remain unexplored. Leaders need a better account of what the work was intended to establish and what it still cannot support.
I want our commercial promises to leave room for an unwelcome test result. That requires more than asking engineers to be brave. The company must fund the investigation, explain what was not examined and accept a commercial consequence when the evidence does not support the requested conclusion. As CEO, I am responsible for making that possible in the way we sell, staff and govern the work.
The journal
Further reading.
02
Capability investment and operating plans5 min read
Recruitment, learning completion and readiness for client responsibility are separate stages. A credible capacity plan funds the senior attention between them and reports progress without turning every starter into an immediately billable resource.
A time-based contract can make useful automation commercially awkward. Leaders should expose the savings, identify the continuing obligations and negotiate a model that funds those obligations without concealing the improvement.
Company colleagues
620
March 2026; three Indian delivery cities
FY 2026 revenue
₹156 Cr
Company revenue; ₹18 crore operating profit
Client organisations
42
India, UK and UAE; scoped testing services
Keep the difficult finding in the final report
I am Elsa Thiruneli, CEO and executive director of Belsori Software Assurance in Coimbatore. Our six hundred and twenty colleagues examine software for forty-two clients, working from three Indian cities. We supply evidence about defined questions and agreed conditions. We do not take over a client's release decision, and a completed test plan is not a certificate that nothing can go wrong. I want that distinction to be understood before an engagement is sold, while there is still time to agree what useful work looks like.
I began by testing industrial instruments, where a result depended on the conditions under which it had been obtained. Retail software introduced me to a different kind of uncertainty: the behaviour that appears when people use a system in combinations its designers did not anticipate. I moved into specialist testing delivery in 2008, became COO in 2015 and took the CEO role in April 2020. My responsibility now includes the economics and ownership expectations of the whole company. Employees hold a substantial stake, but that does not remove the need to make difficult choices about investment, price and capacity.
Four recent decisions explain my approach more clearly than a list of services. We stopped rewarding raw defect counts. We funded an apprenticeship cohort without assuming immediate billable work. We changed the commercial model for an automated regression service so that clients could share the benefit of fewer execution hours. And we retained a qualified technical conclusion when a client wanted a broader endorsement before paying a milestone. In each case I had to reconcile a sound technical position with the resources and commercial behaviour needed to sustain it.
A smaller count can represent harder work
A nine-engagement review showed that raw finding totals rewarded numerous minor observations while giving little credit to investigating consequential failure paths. We replaced the count-based component of team recognition with agreed risk coverage and retest evidence. The review did not establish a universal defect rate or prove that any release was risk-free.
Learning needs an honest capacity budget
A thirty-six-person apprenticeship cohort received a ₹1.9 crore twelve-month learning and support allocation, with six months of supervised practice. Thirty-one completed the programme; twenty-four had entered scoped billable assignments by the review date. Those figures describe different stages, and neither completion nor placement was presented as a professional certification.
Better automation should not need a hidden invoice
For one recurring regression engagement, repeat execution effort fell from 1,200 to 420 hours per release window. We negotiated a defined-scope service at ₹7.8 lakh per window, down from the earlier ₹9.6 lakh time-based bill. The client saved money and we retained funding for maintenance and reserved review capacity; the agreement did not promise unlimited change.
Current focus
Our next task is to improve how engagement teams explain exclusions and unresolved findings to decision-makers outside engineering. A director or business sponsor should be able to see the conditions examined, the areas still uncertain and the person responsible for the next decision. I am also reviewing how our five executive heads fund specialist depth without relying on permanent overtime or the convenient assumption that every trained colleague is immediately ready for every client assignment.
Conversations I welcome
I would consider one external non-executive role where service quality, technical evidence and commercial incentives need to be examined together. I have one current executive directorship and an indicative allowance of eighteen days annually for a suitable external appointment. Both the actual calendar and permission from my current board would need to be settled before I accept.
Beyond the page
The record behind the perspective.
In the boardroom
Belsori Software AssuranceExecutive director · April 2020–present
Formation
Bachelor of Engineering, ElectronicsThesvani Engineering College · 1994
Working with Elsa on the apprenticeship investment since 2023, I have seen her distinguish the people we recruited from the capability we could responsibly commit to clients. She made us fund the learning rather than hide its cost in a staffing forecast.
Pranav LershaniFinance director, Belsori Software Assurance; executive colleague since 2023The full record ↗
Contact
I would consider one external non-executive role where service quality, technical evidence and commercial incentives need to be examined together. I have one current executive directorship and an indicative allowance of eighteen days annually for a suitable external appointment. Both the actual calendar and permission from my current board would need to be settled before I accept.
