Linkrunner: brand, dashboard, web, and the experiments around it

TL;DR

I'm the founding design engineer at Linkrunner, India's first AI-native Mobile Measurement Partner (MMP). My remit started as "make the brand and the dashboard not look like every other ad-attribution tool," and grew into owning every surface customers and prospects touch, plus running the 0→1 experiments around what we sell next.

What an MMP is, in one line

A Mobile Measurement Partner sits between mobile apps and ad networks and answers which ad got us this user. Most MMPs are dashboard-heavy enterprise tools that look like Bloomberg terminals. We're rebuilding that category around AI-native primitives instead of dashboards.

What I own

Brand. Identity, type, color, motion. Not just the logo, the visual grammar that has to hold up across the dashboard, the marketing site, partner decks, ad creative, and (now) the AI product surfaces.

Dashboard. The product surface our customers actually live in. Designing for analytics + attribution data without falling into the "table on a table" trap most MMPs default to.

Web. I built the original linkrunner.io end-to-end in Framer, design + build, no separate dev handoff. Right now I'm rebuilding it in code (Next.js, vibe-coded with Claude Code and Codex) to match the new brand. The beta lives at linkrunner-website.vercel.app, same designer, different toolchain, same person shipping it.

The experiments

Beyond the core surfaces, I run 0→1 product experiments around new monetization routes. The thesis: an AI-native MMP shouldn't only sell attribution, it should sell every adjacent thing that AI now makes cheap to produce.

Creative Intelligence

A platform that ingests your brand, products, website, and app, and then generates marketing creatives on demand, match the brand voice, match the trend, match the platform.

I designed and built the entire thing. It's the first product surface where Linkrunner is the generator, not just the measurer.

Vibe marketing tool (in prototype)

A Slack-native product that turns mention-and-respond into a creative pipeline. Tag @linkrunner in a Slack thread with a prompt, get an ad / creative / social post back in-thread.

The model is closer to how Midjourney grew on Discord than how SaaS dashboards grow. Distribution-first product thinking, every output is also marketing for the tool that produced it.

Product philosophy I'm applying

Reduce friction to zero. No dashboard required to start. No onboarding. Trigger = mention + prompt. The interface is the conversation.

Public output = growth. Every result is visible and shareable. Every interaction is a marketing impression. The tool runs its own growth loop.

Speed over depth (for v1). Instant response matters more than a perfect output. Viral loop > feature completeness. Ship the loop, then improve the model.

AI as a backend primitive. Treat LLMs as infrastructure (orchestrated through Temporal in our stack), not as features bolted onto a dashboard. The product is built on top of AI, not around it.

What this role actually is

The job title says "Founding Design Engineer" but in practice it's product + design + brand + the front half of engineering. I:

• define product behavior, not just screens
• shape growth loops as part of the design
• pick the technical primitives (Framer → Next.js → vibe-coded with Claude Code) that fit the team's velocity
• align AI capability with the UX we actually want

Most days the seam between "designer" and "engineer" doesn't really exist in this role, it's whichever one moves the work forward fastest.

Why I'm sharing this

Most of my older case studies on this site are pre-Linkrunner, Figma-side product design from the freelance era. Linkrunner is where the design-engineer label became literal: I design it, I build it, and I'm accountable for whether it ships and converts.

If you're trying to figure out what working with me looks like in 2026, this is the most current data point.