Back to projects

WhatsApp to Video

A mobile app that turns conversations into video stories

KMP SwiftUI Kotlin Remotion AI
Role Solo Developer
Timeline 2024 - Present
Platform iOS, Android (KMP), Web

The idea

Group chats hold some of the funniest, most emotional, and most memorable moments of our lives. But they stay buried in scroll, forgotten between thousands of messages. WhatsApp to Video changes that. It transforms any WhatsApp chat export into a polished, shareable video story that captures the highlights of a conversation.

How it works

The flow is simple. Users export a chat from WhatsApp and open it in the mobile app. The parser extracts messages, timestamps, media references, and participant data. Then the AI layer kicks in, analyzing the conversation to identify key moments, emotional arcs, and natural story beats. It generates narration that ties the highlights together into a coherent narrative.

The rendering engine, built with Remotion, produces a video with animated chat bubbles, smooth transitions, and styled typography. Users can customize the visual theme, pick which moments to include, and adjust the pacing before the final render.

Chat parsing and preview

Mobile-first with KMP

The mobile apps are the primary experience. Kotlin Multiplatform (KMP) powers the shared business logic between iOS and Android. The chat parser, data models, and AI prompt orchestration live in the shared module, which means both platforms behave identically without duplicating code. iOS uses SwiftUI and Android uses Jetpack Compose for fully native UI on each platform.

Web companion

A secondary web version is built with Next.js and deployed on Cloudflare Workers for edge performance. Video rendering is handled server-side with Remotion, producing high-quality output that users can download or share directly to social media.

Video rendering pipeline

AI-powered narration

The AI does more than summarize. It understands context, detects humor, picks up on recurring inside jokes, and identifies emotional turning points. The generated narration feels natural, like a friend retelling the story of the chat. Users can choose between different narration styles, from cinematic documentary to casual recap.

Key technical decisions

  • KMP for shared parsing and business logic across iOS (SwiftUI) and Android (Kotlin)
  • Remotion for programmatic video rendering on the server side
  • Cloudflare Workers for low-latency uploads and edge processing on the web
  • AI prompt chaining to extract highlights, generate narration, and suggest visual pacing
  • Incremental rendering so users can preview before committing to a full export
Next project Foot Evo Agent