# Products

Two locally-deployed AI products: Reception, an AI front desk that answers calls and web inquiries 24/7; and Advisor, a customer-service assistant that augments your human agents in real time. Both keep your data on-premise.

---

# Products

Two systems are in production today, both built from our atomic capabilities and both
deployed locally on your infrastructure.

## Reception — AI front desk

**Reception** is an AI receptionist that answers phone calls and web inquiries around
the clock. It books appointments, triages requests, and routes callers to the right
person — so nothing is missed after hours, and no call recording is ever sent to a
cloud. It speaks English, Mandarin, and Cantonese.

## Advisor — customer service assistant

**Advisor** keeps your firm's customer-service expertise **inside the company**. The
knowledge of your best reps is captured rather than lost when they leave, and served
to every agent in real time — so a newer rep reaches veteran performance with far
less training. While a human agent handles the conversation, Advisor grounds answers
in your own knowledge base and surfaces the right clause, answer, and suggested next
action. **Multi-tier permissions** also let it back up your senior professionals
directly, giving each role only the information and actions appropriate to their
authority. Every interaction stays on-premise.

## Built from the same stack

Both products draw on the same layers: a **locally-deployed LLM**, the **speech
pipeline**, and **retrieval (RAG with rerank and GraphRAG)**. Because we control
those layers, we can tune either product to your domain and recombine them into a
custom agent when you need something different.