Rahul ZeqvethiChief Data Officer · Essays & perspective
Four heads within the insurer
43 data colleagues
₹8.6 crore FY2026–27 expense; ₹2.4 crore separate improvement programme
Field statistics to enterprise data leadership
29 years
Five contiguous appointments from July 1997
Company context
19 branches
920 employees and ₹1,860 crore FY26 gross written premium; not personal authority

Both service rates improved while the total fell

I am Rahul Zeqvethi, chief data officer at the fictional Talqeri General Insurance in Mangaluru. One report I approved showed a declining overall callback rate. The arithmetic was right; the account was incomplete. Digital callbacks improved from 80% to 90%, and assisted callbacks from 60% to 65%. The business handled a much larger share of assisted requests in the second period. The combined rate fell even though both channel rates rose. I had allowed one direction to stand for a more complicated service change.

I grew up in Kanpur and began as a field statistician in July 1997. Sampling and research taught me to ask who entered a calculation and who remained outside it. Insurance added decisions whose consequences required those questions to be answered consistently, not just for a research assignment. I now lead forty-three data employees through four heads, with ₹8.6 crore FY2026–27 operating expense and a separate ₹2.4 crore data-improvement programme. Talqeri's ₹1,860 crore FY26 gross written premium and 920 employees describe the company, not my delegated authority.

My work concerns the explanation that travels with a number. We revised the callback account, kept missing turnover separate from a declared zero, reconsidered an aggregate release whose totals exposed hidden cells and declined an external dataset whose usable contribution did not support its proposed cost. Each case retained an uncomfortable limit. Better conditional callback rates did not erase missed promises. A completed field did not prove a declaration true. A restricted table did not establish universal confidentiality, and declining a purchase did not prove the supplier's records were worthless.

The changing mix belonged beside the headline

Digital callbacks were 800/1,000 then 450/500; assisted callbacks were 180/300 then 650/1,000. The combined result moved from 980/1,300 to 1,100/1,500. A ₹52 lakh reconstruction retained all denominators.

Unknown turnover stayed unknown

Of 2,400 reviewed files, 420 had missing turnover, ninety declared zero and 1,890 stated a positive amount. A ₹48 lakh confirmation exercise resolved 312 missing entries and left 108 unresolved.

More purchased fields did not establish more useful decisions

A ₹1.6 crore annual data proposal was declined after a ₹27 lakh evaluation. Of 480 reviewed records, 72 offered a newly supplied field and only 36 met the agreed current-use criteria.

Current focus

Making insurance management measures useful without allowing aggregation, completeness or volume to imply more than the evidence supports.

Conversations I welcome

One first board discussion concerning data judgement and management evidence, after preparation, employer consent and appointment-specific review.

Beyond the page

The record behind the perspective.

Formation

  • Bachelor of Science in StatisticsVorzemi Statistical College · 1997
Rahul recommended declining the data contract without turning an unapproved annual proposal into a claimed saving. He kept the evaluation expense and the remaining stewardship need visible, which made the recommendation useful for the next decision as well as this one.
Saira DazquriCEO, Talqeri General Insurance; fictional colleague
The full record ↗

Contact

One first board discussion concerning data judgement and management evidence, after preparation, employer consent and appointment-specific review.

  • Board appointments
Write to Rahul Contact page