Software Engineer III · Bangalore, India

Riya Sharma

Software engineer shipping across the stack — web, mobile, and the AI systems in between. These days that means being a core engineer on RapidNative, turning prompts into real mobile apps; before that, design systems and GlueStack UI at GeekyAnts. I write about the engineering behind it.

4+ yrs shipping web & mobile · 1,100+ commits on RapidNative · GlueStack UI contributor

## featured work

RapidNative

The fast, collaborative AI-native mobile app builder. Real mobile apps from a prompt — native device capabilities, real-time collaboration, clean Expo code export, and publishing to the App Store and Play Store.

Next.jsTypeScriptExpoSupabase

reactnative-run

In-browser React Native bundler. I contributed ESM/Metro compatibility fixes, web shims for native-only packages, and runtime error-handling semantics.

TypeScriptMetroesbuild

vibecode-db

A Supabase-1:1 database client with swappable mock and real adapters, so generated apps render seeded data instantly and swap to a live backend without code changes.

TypeScriptSupabasePostgREST

GlueStack UI

Open-source universal component library for React & React Native (5,000+ GitHub stars) — helped build components and frontend architecture during my time at GeekyAnts.

React NativeTypeScriptDesign Systems

Aqua Delivery Hub

Side project: a water-delivery platform with three role-based portals — customers ordering refills, delivery partners managing orders and routes, and an admin dashboard overseeing orders and partners.

ReactViteFirebaseshadcn/uiTailwind CSS

## latest writing

all posts →

## about

I started as an intern at GeekyAnts in 2022 and grew into a Software Engineer III in under four years — going from building design systems and helping build GlueStack UI (open source) to now being a core engineer on RapidNative, where my code runs every time someone turns a prompt into a working mobile app.

What gets me excited is the layer most people never see: how a React Native bundle runs inside a browser, how an LLM agent decides what code to write, how a billing system stays honest. I like owning problems end to end — the feature, the infra under it, and the edge cases nobody wrote down.

Off the clock, I'm usually building something anyway — side projects like a water-delivery platform, or pushing new AI coding tools to their limits to see where they actually break. Punjab raised me, Bangalore keeps me busy.

## what i do

AI-powered products

LLM pipelines end to end — multi-provider routing, agent behavior, prompt-to-app code generation, and metering what every generation costs.

React Native, deeply

Beyond screens: in-browser bundlers, Metro internals, web shims for native-only packages, and Expo tooling that ships real APKs.

Platform engineering

The unglamorous parts that make products trustworthy — billing and credits, preview domains, deployment pipelines, and design systems.

## toolbox

Frontend & Mobile

JavaScript (ES6+)TypeScriptReactReact NativeExpoNext.jsReact HooksTanStack Query (React Query)ReduxTailwind CSS

AI & Tooling

LLM API integrationAI agents & prompt engineeringClaude CodeCursor

Backend & Data

Node.jsExpressLaravelPHPREST & GraphQLAuthenticationPostgreSQLMySQLSupabaseFirebase

Developer Tools

GitGitHub ActionsPlaywrightDockerVercelCloudflareCI/CD & deployment pipelinesStorybookBundlers (Metro, esbuild)Design systemsBilling systemsPerformance optimization

edu: B.E. Computer Science Engineering · Chitkara University · 2023

## experience

Software Engineer III · Shaper Studio IncOct 2025 - Present

Core engineer on RapidNative, the AI-native mobile app builder: describe an app and get a real React Native app, ready to publish to the App Store and Play Store. 30,000+ users, 39,000+ apps generated, 1,100+ commits across the platform, bundler, and infrastructure.

  • Designed scalable frontend architecture and led technical design discussions as one of RapidNative's core engineers.
  • Built major parts of the billing and credits system serving hundreds of paying teams — 140,000+ AI generations metered.
  • Cut AI generation time-to-first-output from ~2 minutes to 30-40 seconds by integrating LLM providers directly instead of routing through OpenRouter.
  • Defined engineering standards for agent conventions, maintained vibecode-db and the Expo base template behind every generated app, and mentored engineers on AI-assisted workflows.
TypeScriptNext.jsReact NativeExpoNode.jsBunSupabaseMetroLLM APIsRedux
Software Engineer II · GeekyAnts India Pvt LtdAug 2024 - Sep 2025
  • Architected and contributed to Shaper Studio, an AI-powered platform for automated app generation and workflow optimization.
  • Designed and implemented AI-driven features via APIs, and built scalable frontend applications with Next.js and React.
  • Optimized global state management with Redux, cutting unnecessary re-renders, and strengthened frontend–backend integration for AI workflows.
Next.jsReactTypeScriptReduxNode.js
Software Engineer · GeekyAnts India Pvt LtdAug 2023 - Jul 2024
  • Developed and scaled reusable design systems for Applaunchkit.
  • Helped build GlueStack UI, GeekyAnts' open-source React Native component library (5,000+ GitHub stars), working on frontend architecture and component libraries.
  • Refactored and optimized legacy frontend systems and improved code quality through code reviews and shared component standards.
ReactReact NativeTypeScriptDesign Systems
Software Developer Intern · GeekyAnts India Pvt LtdMay 2022 - Jul 2023
  • Built a Spotify-inspired full-stack application with React and Firebase — authentication, real-time data, and dynamic UI components.
  • Implemented backend features using Laravel, PHP, and REST APIs.
ReactFirebaseLaravelPHP

Let's build something.

Open to interesting problems in developer tools, React Native, and AI-powered products. The fastest way to reach me is email.