[KongchangAI]
Product

Unity Gateway

A gateway product providing AI agent call-chain tracing capabilities. It is used to monitor and analyze complete data across agent invocation chains, supporting observability construction for AI infrastructure.

Core Facts

Timeline (last 90 days)

Oct 8

Unity Gateway is an extension component of Unity Catalog, Databricks' unified data governance system.

Unverified50%
Oct 8

Databricks provides Unity Gateway, which is responsible for uniformly handling governance, monitoring, and security at the invocation layer.

Unverified50%
Oct 1

Unity Gateway is responsible for three core functions: selecting the right model for the right task, governing every call, and controlling costs across models.

Unverified50%
Sep 29

The name Unity Gateway implies integration with Unity Catalog, Databricks' data and AI asset governance platform

Unverified50%
Sep 29

Through a unified gateway, when a new model becomes available, engineering teams only need to complete configuration within the gateway to make it available to all employees, without having to modify each downstream application individually

Unverified50%
Sep 29

Unity Gateway supports setting model access permissions by department, logging call inputs and outputs to meet audit requirements, and tracking inference costs by project.

Unverified50%
Sep 29

Claude Sonnet 5.5 on Databricks undergoes unified governance through Unity Gateway.

Unverified50%
Sep 29

At a webinar hosted by Matei Zaharia and Kasey Uhlenhuth, the two demonstrated two toolsets: Unity Gateway and Omnigent.

Expired50%
Sep 29

Unity Gateway can proactively send notifications when usage approaches or exceeds budget.

Unverified50%
Sep 29

Unity Gateway provides teams with a unified view across models, agents, and workflows, consolidating dispersed AI calls into a single observable plane.

Unverified50%

13 more timeline events

All Facts (20)

Verified

Databricks achieves unified governance over model access through Unity Gateway.

85%
Verified

Unity MCP adopts a client-server architecture consisting of three layers: the Unity-side plugin, the MCP server, and the AI client connection layer.

80%
Unverified

Unity Gateway is positioned as a unified AI gateway.

70%
Unverified

Every call to GPT-6 Astra flows through Databricks' Unity Gateway, benefiting from unified governance, audit trails, and cost controls.

70%
Unverified

Unity Gateway is an extension component of Unity Catalog, Databricks' unified data governance system.

50%
Unverified

Databricks provides Unity Gateway, which is responsible for uniformly handling governance, monitoring, and security at the invocation layer.

50%
Unverified

Unity Gateway is responsible for three core functions: selecting the right model for the right task, governing every call, and controlling costs across models.

50%
Unverified

Through a unified gateway, when a new model becomes available, engineering teams only need to complete configuration within the gateway to make it available to all employees, without having to modify each downstream application individually

50%
Unverified

The name Unity Gateway implies integration with Unity Catalog, Databricks' data and AI asset governance platform

50%
Unverified

Unity Gateway supports setting model access permissions by department, logging call inputs and outputs to meet audit requirements, and tracking inference costs by project.

50%
Unverified

Claude Sonnet 5.5 on Databricks undergoes unified governance through Unity Gateway.

50%
Unverified

Unity Gateway provides teams with a unified view across models, agents, and workflows, consolidating dispersed AI calls into a single observable plane.

50%
Unverified

Unity Gateway can proactively send notifications when usage approaches or exceeds budget.

50%
Unverified

Unity Gateway goes further than a unified gateway by incorporating enterprise-grade permission management and quota management.

50%
Unverified

The naming of Unity Gateway continues the brand lineage of Databricks' previously launched Unity Catalog, which is its unified catalog and governance service for data and AI assets

50%
Unverified

The principle of selecting the right model for the right task is moving from a concept toward an engineerably deployable capability, and cost governance should be incorporated into consideration at the architecture design stage

50%
Unverified

Genie One MCP provides fine-grained policy control and audit logs through Unity Gateway, enabling centralized governance

50%
Unverified

Concurrence uses Unity Gateway to govern clinical AI at the trillion-token scale.

50%
Unverified

Distribution through the Unity Gateway means enterprise customers can invoke GPT-6 Astra within Databricks' unified data and governance framework without separately integrating OpenAI's API.

50%
Unverified

The case used Unity Gateway's trace tracking and Genie One's analytical capabilities as the core tool combination.

50%

Source Articles