Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.osvi.ai/llms.txt

Use this file to discover all available pages before exploring further.

What is OSVI?

OSVI is an AI-powered voice and communication platform. It lets you deploy AI agents that can hold real phone conversations and chat sessions with your customers — handling sales, support, scheduling, and more without human involvement.

Core Concepts

Agents

An agent is the core unit of OSVI. It defines:
  • Personality and behaviour — through a system prompt
  • Voice — via a TTS (text-to-speech) configuration
  • Language understanding — via an STT (speech-to-text) configuration
  • Intelligence — via an LLM (large language model) configuration
  • Tools — functions the agent can call during a conversation (e.g. transfer to a human, look up a record)
There are two agent types:
TypeDescription
Single-promptOne system prompt governs the entire conversation. Best for straightforward, linear workflows.
Multi-promptThe conversation is broken into named states, each with its own prompt and transitions. Best for complex, branching flows.

Calls

An outbound call is initiated via the POST /call API. OSVI dials the number, connects the agent, and runs the conversation. When the call ends, OSVI posts the result to your webhook_url.

Chat Sessions

A chat session is a text-based conversation with an agent. You start a session, exchange messages, and close it when done. Chat sessions are stateful — the agent remembers the full conversation history within a session.

Campaigns

A campaign is a batch calling operation. You upload a CSV of contacts, assign an agent, and OSVI works through the list — handling retries, scheduling, and queue management automatically.

Knowledge Base (RAG)

Agents can be connected to a knowledge base — a set of documents that the agent can search during a conversation to answer questions accurately. Documents are stored as vector embeddings and retrieved via semantic search.

Typical Integration Patterns

Outbound calling campaign

Upload a contact list to a campaign. OSVI calls each number, runs the agent conversation, and posts results to your webhook.

Inbound support chat

Embed a chat widget on your site. Use the Chat Session API to start sessions and exchange messages in real time.

Appointment reminders

Trigger POST /call for each appointment. Pass appointment details via additional_data so the agent can reference them in conversation.

Lead qualification

Run agents on inbound leads. Collect structured answers via post-call analysis and push results to your CRM via webhook.