AI Voice Soft Skills Coach
Budget: $15 – $25 AUD
Project Title:
Juke POC Development – AI Roleplay Platform for Solo & Multi-User Modes
Overview:
We’re building a proof-of-concept AI training tool that helps users practice soft skills (communication, leadership, conflict resolution) through realistic, voice-driven roleplay.
POC Features:
1. Mode Selection – Solo (1 human + AI) or Multi-User (2–4 humans + AI facilitator).
2. Participant Setup – Basic profile info (job, industry, email).
3. Roleplay Sessions – Real-time conversations up to 10 minutes, scenario setup & AI role assignment.
4. AI Interaction – OpenAI Realtime API for STT/TTS, GPT-4 for adaptive responses.
5. Feedback Reports – AI-generated reports (individual + team), viewable in UI and exportable to PDF.
6. AI Verbal Feedback – Spoken session summary (optional).
7. Admin Panel – CRUD for prompts (scenario, roleplay, reporting).
Technical Preferences (open to alternatives):
• Frontend: React (Next.js)
• Backend: Node.js
• Audio/Video: WebRTC (OpenAI Realtime API, or LiveKit if needed)
• AI: GPT-4, OpenAI Realtime API (STT/TTS)
• DB: MongoDB
Deliverables:
• End-to-end working demo (solo + multi-user roleplay).
• At least one demo scenario track (e.g. Negotiation & Influence).
• Feedback report (on-screen + PDF via email).
• Basic admin functions.
• Works in Chrome, Safari, Firefox (desktop priority, mobile/tablet if feasible).
• Performance metrics (latency, STT/TTS accuracy).
What We’re Looking For:
• Proven experience with voice/video AI tools (WebRTC, STT/TTS).
• Strong skills in NLP/LLM integration (GPT-4 or equivalent).
• Experience building interactive training apps or roleplay bots.
Please share examples of similar projects with your bid
Juke POC Development – AI Roleplay Platform for Solo & Multi-User Modes
Overview:
We’re building a proof-of-concept AI training tool that helps users practice soft skills (communication, leadership, conflict resolution) through realistic, voice-driven roleplay.
POC Features:
1. Mode Selection – Solo (1 human + AI) or Multi-User (2–4 humans + AI facilitator).
2. Participant Setup – Basic profile info (job, industry, email).
3. Roleplay Sessions – Real-time conversations up to 10 minutes, scenario setup & AI role assignment.
4. AI Interaction – OpenAI Realtime API for STT/TTS, GPT-4 for adaptive responses.
5. Feedback Reports – AI-generated reports (individual + team), viewable in UI and exportable to PDF.
6. AI Verbal Feedback – Spoken session summary (optional).
7. Admin Panel – CRUD for prompts (scenario, roleplay, reporting).
Technical Preferences (open to alternatives):
• Frontend: React (Next.js)
• Backend: Node.js
• Audio/Video: WebRTC (OpenAI Realtime API, or LiveKit if needed)
• AI: GPT-4, OpenAI Realtime API (STT/TTS)
• DB: MongoDB
Deliverables:
• End-to-end working demo (solo + multi-user roleplay).
• At least one demo scenario track (e.g. Negotiation & Influence).
• Feedback report (on-screen + PDF via email).
• Basic admin functions.
• Works in Chrome, Safari, Firefox (desktop priority, mobile/tablet if feasible).
• Performance metrics (latency, STT/TTS accuracy).
What We’re Looking For:
• Proven experience with voice/video AI tools (WebRTC, STT/TTS).
• Strong skills in NLP/LLM integration (GPT-4 or equivalent).
• Experience building interactive training apps or roleplay bots.
Please share examples of similar projects with your bid