I will build a multimodal ai voice assistant with android control


About this gig
A production voice assistant, built end-to-end - not a demo.
What you get:
- Natural voice conversation: speech-to-text input, spoken answers
- Document and photo understanding: PDF, Word, Excel, PPT and images
- Web-connected answers with cited sources
- Streaming chat UI and background task handoff between phone and web
- On request: Android phone control via Accessibility APIs (open apps, tap, read screen, send messages, set alarms)
What is NOT included: my AIDA product itself - its brand, app-store listings, hosted backend and prompt library. You get a custom assistant app built for your use case, plus the source code of that build.
How I build it: LLM orchestration with a tool-calling pipeline, optional local ASR for privacy, clean mobile UI and a deployment guide. Every deliverable ships with setup instructions and a testable demo.
17 years in AI R&D; my voice assistant product has 16,000+ downloads on Android, iOS and the web.
Message me with your use case and preferred LLM provider before ordering.
Get to know Arthur Wu
- FromChina
- Member sinceApr 2025
- Avg. response time3 hours
- Last deliveryabout 6 hours
Languages
English, Chinese
Other AI Development Services I Offer
FAQ
Can the assistant work without cloud APIs?
Yes. I can ship a fully local setup - Whisper for speech, a local LLM (Llama/Qwen) served with vLLM - so no audio or text leaves the device. Cloud APIs are faster and cheaper to start with; I will recommend the right mix for your case.
