Product demo
AI Voice Contact Center
A voice agent for CoveScreen that listens, thinks, and answers from verified CoveScreen knowledge — today running entirely on the local engine, with the human handoff openly simulated.
You speak
Your browser runs the built-in Web Speech API — no account, no API key, nothing leaves the page for transcription.
The line answers
The server retrieves relevant facts from the CoveScreen knowledge base (FAQs + documents) and the local engine drafts a short, spoken reply.
Knows its limits
When there is no confident answer, the agent says so honestly — and can offer a clearly labelled demo handoff (no real call is ever placed).
What the demo builds on
Curated, verified CoveScreen knowledge: how we work, what we deliver, and how to reach the team. The agent only ever answers from that content — it never invents services, prices, statistics, or policies.
- AI automation & process intelligence
- Cybersecurity operations
- Cloud infrastructure & DevOps
- Custom software & mobile apps
- Dashboards & analytics
- Digital consultancy
Honest by design
This first build is a fully local demo engine: speech is processed by the browser, retrieval runs on the server, and no external AI keys are involved. The path to a production-grade line is documented but not activated here.
- Engines: browser Web Speech API for STT/TTS; local retrieval-based language engine.
- Handoff: simulated. The line records the request clearly, but never connects a real agent or places a call.
- Telephony: Asterisk/SIP is a future integration point, not an active dialer.
Integration path for real deployments: talk to the team.
Staff sign in to manage the voice line.