From sitemap to launch,
in 14 days.
Bolna.ai is a Voice AI platform built for India. Ahead of their Y Combinator announcement, they had two weeks to replace a website that hadn't kept up with the company. I led the redesign end to end, starting with structure, not screens.
Two weeks wasn't a design problem.
It was a prioritisation problem. With timelines that tight, every decision either creates momentum or quietly costs it.
The messaging felt outdated. The structure couldn't carry a growing product ecosystem. The visual language no longer reflected the credibility of a YC-backed company about to step in front of a much larger audience.
Inside a short window, I had to understand the Voice AI space, learn the product, align with stakeholders, and ship a site that worked for three different visitors at once.
- 01
Enterprise customers
Looking for credibility, scale signals, and proof the product can carry their use case.
- 02
Developers
Wanting documentation, APIs, and a fast route to the technical detail.
- 03
Casual visitors
Curious about Voice AI in India, needing a clear, plain-language entry point.
Starting with the sitemap, not the pixels.
The first thing I pushed back on was jumping straight into screens. Two parallel pieces had to happen before Figma made sense: defining the structure everything else would hang from, and understanding the field Bolna was about to step into.
The sitemap first.
What pages existed, how users moved between them, what content deserved prominence, what could live deeper. Getting that structure right early gave the design work a frame to push against, and meant no painful reshuffling once we hit high-fidelity.
Then the landscape.
A fast review of Voice AI sites, both global names and Indian players. The question wasn't who looked best. It was where Bolna should align with familiar industry patterns, and where it had a real opportunity to stand apart.
Fast feedback loops, not perfect deliverables.
With two weeks to ship, I optimised for momentum. Low-fidelity wireframes built for discussion, not polish, went in front of founders, product, and engineering early. By the time high-fidelity designs landed, the core decisions had already been reviewed multiple times. No surprises at the end.
Show something, align on it, keep moving. Done beats perfect when the launch date doesn't move.
A site that reflected the moment.
Best experienced live, or in the embedded Figma prototype below. Designed for desktop, so open it on a larger screen for the intended view.
The kind of project you remember for the people.
A two-week timeline only works when trust is already there. Bolna's team treated me as part of the team from day one, and that made every decision faster.
“When I pushed for decisions, start with the sitemap, simplify scope here, the founders were open, asked thoughtful questions, and trusted the process. That trust is the only reason a two-week launch was possible at all.”
Decisive, open to pushback.
Strategic calls landed in the same conversation they were raised. Nothing got stuck waiting for next week.
Context, not assumptions.
Working alongside the engineers helped me understand the product as it actually worked, so the design didn't promise things it couldn't ship.
Feedback that moved.
Reviews happened in parallel, not in series. Founders, product, and engineering stayed aligned without long redesign cycles.
The site shipped. The thinking stayed.
The website went live in time for the YC announcement, but the part I'm proudest of isn't a visual decision.
The sitemap and competitor work kept guiding the team after the project ended. The foundation outlasted the redesign, which felt like a sign that the thinking behind it was built to last.
Selected work shown here. The full prototype is best experienced live.
Happy to walk through the design decisions, the wireframes, and the parts that didn't make it past day three.