A forecast can become more persuasive as its limitations become less visible. My engineering responsibility is to keep missing inputs, uncertain assumptions and user changes legible in the product. The client should be able to question a result without needing to reverse-engineer how the software arrived at it.
I design treasury software so a useful projection does not acquire the appearance of money already available.
I am CTO of Meralqen Treasury Software in Cuttack. We provide cash-planning and forecasting tools to 58 corporate clients, with 210 employees and ₹62 crore of software and support revenue in FY2025–26. I lead eighty-five engineers through four heads and administer an approved ₹18 crore engineering budget. Our product helps clients examine their own financial assumptions. It does not hold funds, execute payments or decide investments, and I do not describe the value of a client's forecast as assets under our management.
I began as a business-software engineer in 1998 and entered treasury products six years later. The work taught me that a convenient default can carry a financial meaning its developer never intended. A missing opening balance can become a zero; a group hierarchy can be mistaken for permission to see every subsidiary; a projected receipt can look as definite as a reconciled bank entry. These are product decisions as well as programming details. They need a clear owner and a test that reflects how the result will be read.
My recent work has focused on incomplete forecasts, group access and the conditions under which an automated projection should decline to produce a figure. We have accepted additional user steps, narrower initial releases and more visible gaps in the output. Those choices are not a rejection of automation. They are an attempt to make it useful without hiding the business judgement that still belongs to the client. I want a treasury team to see where the software is helping and where its own evidence must do more work.
Missing balances stopped producing a complete total
A review of 146 incoming source sets found fourteen without a current opening balance. We changed the product to mark the affected forecast as incomplete and funded ₹28 lakh of engineering and migration work.
Group structure separated from access
A proposed consolidated view prompted a review of forty-two user permissions across a pilot group. Twelve lacked a documented business justification and were removed before the revised feature was accepted.
A forecast could say evidence was insufficient
In a nine-client pilot, three businesses had changed operating patterns that historical data could not represent reliably. We held automated figures for those cases and retained an explicit client-owned scenario route.
Current focus
I am strengthening the product's explanation of assumptions and manual changes across the forecast lifecycle.
Conversations I welcome
I welcome product-engineering leadership where financial interpretation matters, and a first board conversation about the consequences of software design.
The leadership record
Trusted with consequential decisions.
Formation
Bachelor of Engineering in Computer EngineeringDaruneli Engineering Institute · 1998
Devashish made the release discussion include what the client would infer from the number. His recommendation gave us an alternative with a cost and a clear limit, not simply an engineering objection.
Ranjana TevlariFictional chief executive, Meralqen Treasury SoftwareThe full record ↗
Contact
I welcome product-engineering leadership where financial interpretation matters, and a first board conversation about the consequences of software design.