MII Realtime Call (ARIA)
AI-first call center with realtime voice agents.

At a glance
- Same-sessionHuman handoff without re-queuing
- Speech-to-speech Single realtime model, no ASR-TTS chain
- Grounded RAGAnswers cited from customer knowledge






Overview
A realtime voice agent platform that answers calls, understands intent, and replies in natural speech across Bahasa Indonesia and English.
The live conversational turn runs on a single speech-to-speech model rather than a chained speech-to-text, reasoning, and text-to-speech pipeline, which keeps the conversation responsive. Answers are grounded in approved customer knowledge through hybrid vector and keyword retrieval, and when confidence is low or a case needs human judgment, an operator joins the same live call instead of the caller being transferred or put back in a queue. Each interaction is captured as structured data for follow-up and reporting.
Case study
- 01
The challenge
Contact centers face high call volumes, long wait times, limited after-hours coverage, and service quality that varies across shifts, languages, and agent availability. Routine questions consume human capacity while complex cases still need empathy and judgment.
- 02
Our approach
ARIA runs the live turn on a single speech-to-speech realtime model instead of chaining separate recognition, reasoning, and synthesis stages, removing the pipeline hops that add conversational delay. Retrieval is exposed to the model as a callable tool, so answers are pulled from approved customer knowledge on demand with provenance attached rather than stuffed into a static prompt. When retrieval confidence is insufficient or the case calls for judgment, the platform brings a human operator into the existing call with full context preserved.
- 03
The outcome
A voice agent that picks up immediately, answers from approved knowledge with sources attached, records conversation context, captures follow-up details, and hands the operator a fully contextualised call when a case must be escalated.
- 04
Impact
Faster first response, more consistent service quality, better peak-volume coverage, richer call records, and human agents freed to focus on sensitive or high-value conversations.
Technologies
09- Azure OpenAI
- Azure Speech
- Realtime API
- Azure AI Foundry
- Azure Communication Service
- Azure AI Search
- RAG
- Human
- Handoff
Related work
MII GenAI Chatbot
An enterprise GenAI chat assistant grounded in your knowledge.
- Azure OpenAI
- Azure AI Search
- RAG
- Cosmos DB






