I built the estimation tool I wished I had as a CPWD engineer.

The Problem

I cleared the Indian Engineering Services exam in 2013 and joined CPWD in 2015. Over the next 10+ years, I worked on public infrastructure projects — government buildings, housing, roads, the kind of work that actually gets used by millions of people. The work itself was meaningful.

But the planning side? That was a different story.

Creating a detailed estimate meant hours of combing through a 600-page DSR PDF, manually looking up item codes, copying rates into Excel, building formulas from scratch, and hoping nothing broke when you changed a number three sheets deep.

And then came tender justification.

Justifying a tender meant redoing the entire analysis of rates — for every single item — with current market rates. Hundreds of items, each with its own nested dependencies. In Excel. By hand. A task that should take hours would stretch into weeks, sometimes months. And if you found an error near the end, you'd have to trace it back through layers of formulas held together by hope.

The First Attempt

I tried to fix it inside Excel first. I built custom spreadsheets with lookup formulas, cascading calculations, and macro-based automation. And it worked — as long as I was the one using it.

But the moment I needed a new estimate for a new project, the whole thing fell apart. Formulas broke. Copied sheets carried stale references. The whole system depended on me babysitting it.

I realised I wasn't dealing with a spreadsheet problem. I was dealing with a software problem.

Building Costimator

I'm not a software engineer by training. I learned what I needed to as I went, and brought in help where I had to.

What I did know was how DSR estimation actually works in practice — the dependency trees, the W/X/Y/Z margin calculations, all the things you only learn at 11 PM before a tender deadline. That knowledge was the hard part. The software was the part I could learn.

The first version was rough. Just a web app that could look up DSR items and calculate an analysis of rates without breaking. But it was already faster than anything I'd built in Excel. And it never lost a formula.

That's where the tagline came from: "Accuracy isn't optional." I didn't want accuracy to be something you had to double-check. I wanted it to be the default.

From Personal Tool to Product

I built Costimator for myself. But word got around.

Colleagues started asking about it. Other divisions. Other departments. Engineers who'd been fighting the same Excel battles I had.

What started as a CPWD-only tool is now expanding across multiple Schedules of Rates — Indian Railways, MoRTH, state governments. The problem isn't CPWD-specific. Every department that uses an SoR has the same pain.

Quitting IES

In 2025, I resigned from the IES to build Costimator full-time. Solving this problem part-time wasn't enough.