Databricks Centralizes Coding AI

The rapid proliferation of AI coding assistants, often referred to as coding agents, presents a significant governance challenge for enterprises. Databricks is addressing this sprawl with its new Databricks AI Gateway, a centralized hub designed to manage and secure these tools.

StartupHub data

Companies working on this

Profiles of the companies named in this story, with founding year, headquarters, and a short description from our database.

Databricks

A unified data analytics and AI platform built on the lakehouse architecture.

Founded

2013

Location

San Francisco, United States

Valuation

$190.0B

Anthropic

Anthropic is an AI safety and research company building reliable, interpretable, and steerable AI systems, best known for the Claude family of models.

Founded

2021

Location

San Francisco, California, USA

Valuation

Private / $100B+ est

OpenAI

OpenAI is an AI research and deployment company dedicated to ensuring that artificial general intelligence benefits all of humanity.

Founded

2015

Location

San Francisco, United States

Valuation

Private / $100B+ est

Gemini

A regulated cryptocurrency exchange and custodian focused on security and compliance.

Founded

2014

Location

New York City, United States

Funding

$50M

As software development shifts towards agent-driven workflows, organizations are eager to adopt these productivity boosters. However, granting these agents access to sensitive company data like design documents and customer tickets introduces substantial security and cost risks. The core problem is ensuring these powerful tools are used responsibly without stifling innovation.

The Coding Agent Sprawl Problem

The AI landscape is evolving at breakneck speed, with new models and coding tools emerging weekly. Developers naturally want the flexibility to use multiple tools, Cursor, Codex, Claude Code, and others, often simultaneously. This adoption, while beneficial for productivity, creates a complex environment for administrators.

Security reviews for each new tool can create bottlenecks. Furthermore, coding agents often require elevated privileges to access critical internal data, raising concerns about unauthorized access. This necessitates robust auditing and governance mechanisms for agent data interactions.

The cost of AI usage is also escalating, becoming a major R&D expense. Balancing developer choice with effective cost guardrails is paramount. Without clear visibility into who is using what, and at what expense, controlling budgets becomes nearly impossible.

Executives lack a unified view of AI tool adoption across the organization. This visibility gap hinders strategic decision-making and the identification of adoption blockers.

Introducing Databricks AI Gateway

Databricks AI Gateway aims to provide developers with tool freedom and administrators with unified governance. It unifies access controls, usage statistics, observability, cost management, and inference capacity into a single platform, offering centralized command over AI agents.

The gateway focuses on three key pillars: Centralized Security and Audit, Single Bill and Cost Limits, and Full Observability within the Data Lakehouse.

Centralized Security and Audit

All agent data access is governed centrally. Audit logs are integrated with Unity Catalog, while MCP servers are managed within Databricks. Centralized tracing is provided via MLflow, ensuring comprehensive oversight of agent activities and data access.

Simplified Cost Management

Admins can set unified cost limits across all tools developers use. Databricks' Foundation Model API supports inference for major models like OpenAI, Anthropic, and Gemini, as well as open-source options. This consolidates billing from various AI services into a single invoice from Databricks, simplifying financial tracking and control.

Developers benefit from day-one support for frontier LLM models, ensuring they always have access to the latest capabilities. This centralized approach eliminates the need for admins to juggle multiple consoles for rate limits and budgets, offering a single budget for developers to utilize across their chosen tools.

Unified Observability

Usage data from AI coding tools is treated as a first-class citizen within Unity Catalog, alongside existing enterprise datasets. This provides a governed framework for deep operational intelligence, making all coding activity auditable and secure. Metrics like lines of code written per user, cost per month per user, and adoption rates are automatically ingested into the Data Lakehouse.

This unified data enables organizations to track adoption by department or region, quantify developer velocity, and proactively plan capacity by monitoring rate limit usage. The platform ensures secure and compliant development workflows, with private data remaining within the Databricks security perimeter and audit-ready logging for compliance reviews.

Databricks is also tackling agentic AI risks with its enhanced security features. The introduction of the Databricks AI Gateway is a significant step towards enabling widespread adoption of AI coding tools while maintaining essential control.

The AI Gateway for coding tools is now available to all Databricks customers, with immediate support for Cursor, Gemini CLI, and Codex CLI.