Learning what the observation could actually answer

The vehicle trace was the easiest evidence to obtain, so it became the evidence we asked to do too much. I had endorsed a collection-completion view based on a vehicle entering a defined proximity around a scheduled point. Nusrat Nezqovi asked what the trace actually established. A selected review of 160 events supported observed collection for 112, passage without collection for 28 and an unresolved account for 20. The location record had not become false. My interpretation had turned presence into a service claim without enough support. We withdrew that claim and retained the distinctions the review could justify.

I grew up in Shillong and completed statistics at the fictional Peqzori Statistics College in June 1997. In July I joined Tezqari Field Research as a field-study analyst. The firm had 48 employees and ₹7.6 crore revenue for FY ending March 1998. I worked in a five-person study group without direct reports or spending authority. Much of the early work concerned how observations were obtained: which locations had been reached, who could answer and what a field code meant. I learned that an apparently complete table could still contain an incomplete account of the people or activity it was meant to describe.

The study-manager role began in April 2004. Tezqari employed 122 people and reported ₹25 crore revenue for FY ending March 2005. Eight colleagues worked through two leads within ₹62 lakh annual expense. I became responsible for explaining study limitations to sponsors who wanted a prompt decision. The issue was rarely whether a percentage could be calculated. It was whether the denominator matched the question being asked. I learned to specify the intended use before collecting the observations, while keeping enough flexibility to record why a planned visit could not be completed.

I moved to Jevqori in December 2010 as head of service evidence. The contractor then employed 205 people and reported ₹43 crore revenue for FY ending March 2012. Eleven colleagues worked through two leads within ₹1.3 crore annual expense. Joining an operating enterprise made me responsible for the continuing account, not merely a study delivered at its end. Field colleagues knew why a route had changed, why an entry was late or why a vehicle could pass a point without completing the intended work. Their explanations needed a reliable place in the data process instead of being treated as informal commentary outside the official record.

In May 2017 I became director of service analytics, a functional executive title without a board seat. Jevqori had 480 staff and ₹91 crore revenue for FY ending March 2018. Eighteen colleagues worked through three leads within ₹2.9 crore annual expense. The remit connected route observations, recorded service events, received weights and resident contacts. It also brought pressure to compress unlike evidence into one performance view. I wanted the company to use its growing information without allowing the convenience of a combined dashboard to erase the reasons its sources differed.

My CDO appointment began in March 2022. Twenty-nine data and research colleagues now work through four leads with ₹5.8 crore FY 2026–27 expense and a separately approved ₹2.1 crore evidence programme. The company's 740 employees and ₹132 crore FY 26 revenue describe a larger enterprise than my delegation. Raghav Zevqori is CEO; Nusrat leads field assurance. Operations, worker safety, environmental practice, contracts, finance and technology remain separately owned. We provide defined collection assignments in seven towns. We do not become a municipal authority because our records help describe a public-facing service.

The route-event correction cost ₹46 lakh and required me to withdraw an interpretation I had endorsed. A later ₹34 lakh comparison retained two area groups whose recorded completion rates rose even while the pooled rate fell. The difference came from changing observation weights, not contradictory arithmetic. We kept the total and the groups together rather than choose whichever account presented the company more favourably. Neither view established household coverage or the causal effect of a particular team.

A ₹41 lakh weight review likewise separated received tonnage from claims about waste generation or household behaviour. Two defined four-week accounts recorded 900 and 1020 tonnes, but the difference did not by itself establish what had changed upstream. A field-study programme then included streets with few or no recorded contacts, instead of allowing complaint volume to determine the whole observation frame. Ten of twelve planned street reviews were complete at the stated cutoff, leaving two postponed. Those unfinished visits remained visible rather than being replaced quietly with easier locations.

I work in Khasi and English, with working Hindi and Telugu. The founder, employee-trust and fund ownership proportions are 43%,27% and 30%; I hold no shares. After twenty-nine years, I am considering a first external board role while preparing for responsibilities beyond my executive function. IICA registration is not yet complete in this fictional record, and no assessment result or exemption is claimed. Twenty-two days are an initial allowance, not a promise independent of employer consent or the real calendar. I would bring a habit of asking what a service measure actually observed, with the memory that my own neatest report once answered more than its evidence could support.

Open to Board appointments

Languages / Khasi · English · Working Hindi · Working Telugu