Wire RN
Most paid installs leak before activation. Wire RN builds a personalised onboarding flow for each user from their install attribution data, then renders it natively on iOS and Android.
The problem
You spent $8 on that install. Your attribution stack (AppsFlyer, Branch, Adjust) knows exactly where the user came from: which ad creative, which campaign, which geo. Then your app opens, and that user sees the same onboarding as everyone else. Same questions, same paywall. None of the signal you paid to collect gets used.
60%+ of paid installs churn before activation. The fix isn't a new design. It's showing each user onboarding that matches why they installed in the first place.
How Wire RN works
Attribution context
AppsFlyer, Branch, or Adjust captures the ad creative, geo, device, and install referrer at first open.
The LLM writes the flow
Wire RN sends the attribution context and your value props to the LLM. The agent returns a structured question sequence as JSON.
A2UI renders native screens
Wire RN reads the JSON schema and renders real native components on iOS and Android. No WebView.
What I built
Wire RN is an open-source SDK that drops into your app. You keep your navigation, your design system, and your analytics. The SDK adds a personalisation layer on top.
Flows that read the install
Wire RN reads the install context (ad creative, geo, device, referrer) and builds an onboarding sequence that matches what brought the user in. The person from your Instagram wellness ad sees a different flow than the one from your Google performance ad.
The A2UI protocol
The LLM returns a typed JSON schema. Wire RN reads it and renders real native components. No WebView, no hybrid. Same schema, same agent, and both iOS and Android render it right.
Drops into your app
Wire RN is a rendering layer. You keep your navigation, your design system, and your analytics. One SDK install, one backend webhook, and your attribution SDK wired in. Under a sprint.
Two live use cases
Built and proven on two real apps: a mental health daily check-in (a 114-second flow the LLM writes) and a food recommendation app (a 76-second personalised flow). Both run on the same SDK.
Two live demos
Both built with Wire RN. Same SDK, different markets, different flows.
Mental health: Self-Mastery
A daily wellness check-in. Wire RN builds the question sequence from the user's goals and past check-ins. The LLM decides which questions show up, in what order, and how they read, per user and per session.
114-second flow · 7 generated screens · live in production
Food recommendation app
A personalised food preferences onboarding. Wire RN asks follow-up questions based on the user's first answers: dietary restrictions, cuisines, cooking time. The flow is different for every user.
76-second flow · 7 generated screens · demo live
Technology stack
Wire RN is live in production on the Self-Mastery app. The LLM writes the daily check-in question sequence from the user's goals and past answers. No two sessions are the same.
Malik Chohra, Founder, CasaInnov
Running a pilot
I'm testing Wire RN with a handful of apps. If you want personalised onboarding for your mobile product, get in touch and we'll build the integration together.