JetBrains Central: The Open Infrastructure Orchestrating the Future of Agentic Development
JetBrains Central: The Open Infrastructure Orchestrating the Future of Agentic Development
JetBrains has unveiled Central, an open infrastructure platform designed to govern, orchestrate, and provide shared context for AI agents across the software delivery lifecycle. By addressing the gap between individual AI coding and organizational orchestration, Central aims to redefine enterprise software development without vendor lock-in.
Artificial intelligence has conquered the code editor, but it has yet to conquer the software delivery pipeline. As of early 2026, the cost of code generation has plummeted, but the complexity of managing autonomous agents across enterprise workflows has skyrocketed.
On March 24, 2026, JetBrains introduced a structural solution to this bottleneck: JetBrains Central. Billed as an open infrastructure for agentic software development, Central serves as a unified control and execution plane designed to govern, orchestrate, and contextualize AI agents working alongside human engineering teams.
Solving the "13% Problem"
The launch of JetBrains Central is directly informed by a stark operational reality. According to the January 2026 JetBrains AI Pulse survey of over 11,000 developers, 90% of developers currently use AI in their daily workflows. However, a mere 13% leverage AI beyond the confines of their individual IDE for organization-level tasks like code review, testing, or infrastructure deployment.
The result is a fragmented development cycle: developers utilize highly capable agents like GitHub Copilot or JetBrains' own Junie to write code rapidly, only to hit a wall of manual friction when deploying, reviewing, and testing that code. JetBrains Central is designed to bridge this gap, transforming discrete AI productivity hacks into a cohesive, enterprise-grade production system.
The Three Pillars of JetBrains Central
To elevate AI from an isolated auto-complete tool to a fully integrated team member, JetBrains Central is architected around three core capabilities,:
- Governance and Control: As agents execute multi-step workflows, organizations require strict oversight. Central provides centralized visibility into agent actions, enforcing security policies, managing identity and access (IAM), and offering granular cost attribution to track token expenditure across teams.
- Execution Infrastructure: Autonomous agents require reliable environments to run, test, and build code. Central provisions cloud agent runtimes, ensuring that agentic workloads do not disrupt local development environments while maintaining high availability.
- Agent Optimization and Context: The most significant hurdle for AI agents is a lack of localized understanding. Central introduces a semantic layer that aggregates organizational knowledge—bridging code repositories, architecture documentation, runtime behaviors, and API schemas. This shared context ensures agents make decisions based on the holistic state of the project, not just a single file.
An Open, Agnostic Ecosystem
Rather than building a walled garden, JetBrains has positioned Central as a strictly open, layered architecture. The platform is designed with a "no vendor lock-in" philosophy, supporting an extensive array of external models and agents.
Development teams can utilize JetBrains Central to orchestrate Anthropic's Claude Agent, OpenAI's Codex, Google's Gemini CLI, or internally custom-built agents. Furthermore, through the Agent Client Protocol (ACP) and Model Context Protocol (MCP), Central seamlessly integrates with both JetBrains and third-party IDEs, as well as essential pipeline tools like Git repositories, CI/CD systems, Slack, Linear, and Atlassian,.
Launch partners for the platform include enterprise AI leaders Anthropic, Google Cloud, and OpenAI, underscoring the industry's push toward interoperable agentic infrastructure.
Shifting Priorities: The End of "Code With Me"
The pivot toward an AI-native production system represents a significant shift in JetBrains' resource allocation. Alongside the announcement of Central, JetBrains confirmed it will retire its "Code With Me" human pair-programming feature.
Version 2026.1 will be the final IDE release to officially support the feature natively. This deliberate sunsetting highlights a transition in enterprise software engineering: the future of collaborative development is increasingly viewed through the lens of human-AI coordination—facilitated by new interfaces like JetBrains Air and Air Team—rather than remote human-to-human terminal sharing,.
Looking Ahead
JetBrains Central will officially open its Early Access Program (EAP) to a select group of design partners in Q2 2026.
As organizations move from experimenting with LLMs to deploying autonomous agents at scale, the demand for enterprise governance and orchestration is paramount. By providing the shared context and guardrails necessary for agents to operate safely in production systems, JetBrains is not just offering a new tool—it is defining the foundational infrastructure for the next era of software engineering.