Perplexity-Style AI Aggregator App
Budget: ₹100 – ₹400 INR
I want to create a cross-platform experience—web, Android and iOS—that feels as fluid and helpful as Perplexity.ai but relies on existing large-language-model APIs rather than an in-house AI engine. Users will be able to choose on the fly whether their query is answered by Claude, Grok, ChatGPT-5 or any other model we later plug in. They can type or speak their question; the app will handle both text and voice input interchangeably and return the response in a clean, citation-style interface.
My priorities are:
• A modern, responsive UI, including follow-up prompts and source snippets.
• Seamless switching between multiple model back-ends via cleanly abstracted API calls, so adding new engines is straightforward.
• High-quality speech-to-text using native OS capabilities or leading cloud services to keep latency low.
• Secure user authentication, basic usage analytics and a settings panel where users can set their default model or toggle voice features.
Acceptance criteria
1. A production-ready build uploaded to the web, Google Play and the App Store.
2. Functional model selector with Claude, Grok and ChatGPT-5 fully integrated; a clear interface for adding future models.
3. Latency for a standard query kept under two seconds for text input on a normal broadband connection.
4. Voice input reliably transcribed in under one second of actual speech time.
5. Clean, documented code and a short README covering environment setup, API keys and deployment steps.
If this aligns with your expertise in multi-model AI integrations and cross-platform mobile/web development, let’s discuss your proposed architecture and timeframe.
My priorities are:
• A modern, responsive UI, including follow-up prompts and source snippets.
• Seamless switching between multiple model back-ends via cleanly abstracted API calls, so adding new engines is straightforward.
• High-quality speech-to-text using native OS capabilities or leading cloud services to keep latency low.
• Secure user authentication, basic usage analytics and a settings panel where users can set their default model or toggle voice features.
Acceptance criteria
1. A production-ready build uploaded to the web, Google Play and the App Store.
2. Functional model selector with Claude, Grok and ChatGPT-5 fully integrated; a clear interface for adding future models.
3. Latency for a standard query kept under two seconds for text input on a normal broadband connection.
4. Voice input reliably transcribed in under one second of actual speech time.
5. Clean, documented code and a short README covering environment setup, API keys and deployment steps.
If this aligns with your expertise in multi-model AI integrations and cross-platform mobile/web development, let’s discuss your proposed architecture and timeframe.