AI Voice Assistant App
Budget: $50 – $0 USD
I’m building a fully-featured voice assistant that users can download on both iOS and Android. Think of it as a pocket-sized helper that can listen, understand, and act on spoken commands—setting reminders, fetching quick answers, controlling smart-home services via API, or triggering in-app actions.
Here’s what I need from you:
• A native or cross-platform mobile app (Swift/SwiftUI + Kotlin, or Flutter / React Native—whichever lets us land on both stores quickly and cleanly).
• On-device and cloud speech recognition so the assistant responds within a second or two, even on slow connections.
• Natural-language understanding: basic intent detection, entity extraction, and a fallback to a large-language-model endpoint when the user asks something outside our hard-coded skills.
• A simple, voice-first UI: press-and-hold microphone button, animated waveform, and a conversational transcript that scrolls.
• Hooks to add new “skills” by hitting REST endpoints—no hard recompiles each time we extend capabilities.
• Deployment-ready builds and a README showing how to run, test, and ship to TestFlight and Google Play Internal Testing.
Acceptance criteria (to keep us both clear):
1. User launches the app, taps the mic, speaks a command, and gets a spoken + on-screen reply in under 3 seconds on a typical 4G connection.
2. Both iOS (14+) and Android (8.0+) install packages pass store linting with no critical warnings.
3. Source code is cleanly organised, documented, and pushed to the private repo I’ll share.
4. A one-page PDF or markdown doc summarises architecture choices and how to add new intents.
I’d like to see a working MVP inside three weeks and use the final week for polish and submission. If this sounds like your wheelhouse, let’s get talking right away.
Here’s what I need from you:
• A native or cross-platform mobile app (Swift/SwiftUI + Kotlin, or Flutter / React Native—whichever lets us land on both stores quickly and cleanly).
• On-device and cloud speech recognition so the assistant responds within a second or two, even on slow connections.
• Natural-language understanding: basic intent detection, entity extraction, and a fallback to a large-language-model endpoint when the user asks something outside our hard-coded skills.
• A simple, voice-first UI: press-and-hold microphone button, animated waveform, and a conversational transcript that scrolls.
• Hooks to add new “skills” by hitting REST endpoints—no hard recompiles each time we extend capabilities.
• Deployment-ready builds and a README showing how to run, test, and ship to TestFlight and Google Play Internal Testing.
Acceptance criteria (to keep us both clear):
1. User launches the app, taps the mic, speaks a command, and gets a spoken + on-screen reply in under 3 seconds on a typical 4G connection.
2. Both iOS (14+) and Android (8.0+) install packages pass store linting with no critical warnings.
3. Source code is cleanly organised, documented, and pushed to the private repo I’ll share.
4. A one-page PDF or markdown doc summarises architecture choices and how to add new intents.
I’d like to see a working MVP inside three weeks and use the final week for polish and submission. If this sounds like your wheelhouse, let’s get talking right away.