Ensemble Docs

Getting Started

Build and deploy AI agents with tools, knowledge bases, and workflows.

Ensemble is a platform for building AI agents. Agents reason, call tools, query knowledge bases, execute workflows, and return structured data — all from a single API or embeddable widget.

Platform

Agents

Configure agents with tools, knowledge bases, and LLM models. Supports multi-agent orchestration.

Tools

HTTP APIs, JavaScript, SQL, GraphQL, MongoDB, notifications, transcription, and RAG.

Workflows

Multi-step deterministic processes with branching, loops, and agent integration.

Knowledge Bases

Upload documents for semantic search and retrieval-augmented generation (RAG).

Deployment

Chat Widget

Embed a React chat component in your application. Supports theming and custom styling.

API

REST API for agent invocation, streaming responses, and session management.

Key Features

FeatureDescription
Multi-Agent OrchestrationAgents delegate to sub-agents. Build supervisor patterns or task-based hierarchies.
Rich WidgetsAgents return structured JSON rendered as charts, tables, forms, and custom components.
VersioningDraft and publish agents and tools. Reference by version for stable production deployments.
Environment SyncPromote configurations from Dev → Staging → Prod across tenants.
Bring Your Own ModelUse any OpenAI-compatible endpoint alongside built-in Anthropic, OpenAI, Google, and other providers.
VoiceReal-time voice conversations with agents (Twilio, WebRTC).

Concepts

On this page