Skip to Content
Introduction

Introduction

Welcome to the CorpAI documentation.

CorpAI is a platform for enabling AI in the enterprise through MCP servers  and AI agents. It consists of an AI gateway, control plane, and self-hosted data plane. It provides centralized deployment, RBAC for tools, credential management, audit logging, and more.

The Challenge

Enabling AI across enterprise systems requires solving for:

  • Security: Granting AI access to tools safely (credential management, vetted servers)
  • Access Control: Ensuring only authorized users can use specific tools
  • Compliance: Keeping sensitive data within your infrastructure
  • Observability: Tracking all AI interactions with your systems

The Solution

CorpAI Gateway: A secure proxy that routes AI agents and MCP server calls.

RBAC: Role-based access control for MCP servers and tools. Define policies by user, team, or role.

Self-Hosted Data Plane: All data (documents, conversations, credentials) remains in your AWS account.

Audit Logging: Complete audit trail for all AI interactions.

Features

FeatureDescription
ChatUnified AI access to your organization’s tools and resources.
MCP Server Catalog230+ MCP servers vetted by CorpAI’s team of security experts.
Custom MCP ServersDeploy your own MCP servers for internal tools or custom integrations.
Knowledge BaseUpload documents (PDF, DOCX, TXT) for AI-powered search and Q&A.
PoliciesDefine who can use which tools. Assign by user, team, or role.
Access RequestsEmployees request access, admins approve with full audit trail.
Audit LogsEvery tool call logged with user, timestamp, input, and output.
Usage AnalyticsTrack requests, tool usage, and user activity.

AI Model

CorpAI uses Amazon Bedrock for AI processing. Default model:

  • Gemma 3 12B (google.gemma-3-12b-it)

All AI inference runs in your AWS account via Bedrock.

Quick Start

  1. Provision infrastructure in your AWS account (~45 min)
  2. Add credentials for your integrations
  3. Deploy MCP servers for your tools
  4. Configure policies for your team
  5. Start chatting with your tools

Next

Getting Started

Last updated on