> ## Documentation Index
> Fetch the complete documentation index at: https://docs.coaxial.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Introduction
> Coaxial is an IAM platform for LLMs. Coaxial provides a developer framework that easily integrates access controls for data, models, and functionality of an LLM-powered application.
## Why Coaxial?
At Coaxial, we have a dream of providing every LLM with identity context. Currently, models simply do not care what information they give to who.
With Coaxial, you can automatically control "how a model acts" based on the user it interacts with. Here is how we provide tailored control over LLM functionality and knowledge:
1. **Auth Integrations**: We intake user data from any LLM-powered application (Okta, Auth0, etc.)
2. **Data Integrations**: Coaxial automatically parses index/table data from vector DB's (Pinecone, Chroma, etc.), datalakes (Databricks, Snowflake, etc.), and more. Supports custom data integrations.
3. **Model Integrations**: We intake models from popular sources including OpenAI and Huggingface. Supports custom fine-tuned models as well.
4. **Functions**: Coaxial provides a simple interface to track your LLM functions/external API functionality.
5. **Provisioning/De-provisioning**: The integrations above produce user, data, model, and function objects. Data, model, and function objects are assigned to different users through the Dashboard or SDK.
Developers must use the SDK to check user access to various resources. After implementation, security executives and LLM admins can instantly control LLM behavior.
## Setting up
To start using Coaxial, you need to setup our Developer SDK and Coaxial Dashboard.
Install the Coaxial Python client locally for easy development.
Set up your dashboard account for instant access controls.