Kore.ai unveils platform to tackle enterprise AI sprawl

Kore.ai has launched its Agent Management Platform (AMP), a software layer for governing and monitoring AI agents across large organisations as deployments multiply and become harder to manage.

Many enterprises now run AI agents across business units, development teams, and cloud environments. This spread can leave central technology and risk functions without a clear view of what is running, what it costs, or which controls apply. Industry analysts have described the trend as "AI sprawl".

Gartner has forecast that enterprises will operate thousands of AI agents across business functions within the next few years. The shift raises operational questions about policy enforcement, auditability, performance monitoring, and how to measure business outcomes.

Kore.ai positions AMP as a single control layer across these deployments, describing it as a unified command centre for AI agents and related AI systems inside an enterprise.

Single control layer

AMP is designed to work across different agent frameworks, cloud providers, and development environments. Kore.ai lists support for LangGraph, CrewAI, and AutoGen, as well as Google ADK, AWS AgentCore, Microsoft Foundry, and Salesforce Agentforce, alongside proprietary systems.

The platform consolidates observability, governance controls, performance monitoring, and value measurement into one layer. Kore.ai presents it as a way for organisations to move from decentralised experimentation to more standardised deployment practices.

A key focus is visibility into what agents do and how they behave over time. In many organisations, deployments begin as team-level initiatives in areas such as customer service, IT support, sales operations, finance, and software engineering. Over time, separate toolchains and model choices can emerge, complicating oversight for security teams and internal audit-particularly when agents handle sensitive data or trigger downstream actions in enterprise systems.

Testing and evaluation

Kore.ai highlights an evaluation studio as a core component of AMP, designed to help teams test agent behaviour, workflows, and outcomes before moving agents into production.

The company also emphasises interoperability. Many governance tools are tightly coupled to a single vendor environment, whether a cloud provider, application platform, or model ecosystem. AMP is positioned for mixed environments where organisations use multiple clouds and agent frameworks.

Cost oversight is also becoming a practical concern. AI agents can drive variable usage depending on volumes, model selection, prompts, tool calls, and how often systems are retrained or evaluated. Kore.ai says AMP can help enterprises track performance and costs, apply governance policies consistently, and detect anomalies or drift. It also says organisations can link AI initiatives to measurable business outcomes.

Enterprise oversight

Organisations are formalising AI governance frameworks, often involving the CIO, CISO, compliance teams, and business owners. Agent-based systems add complexity because they can chain actions across multiple tools and datasets. This can make it harder to trace the path from an initial user request to an automated action, especially when multiple models or third-party components are involved.

Kore.ai's strategy targets this operational gap by treating AI agents as assets that need lifecycle management, policy controls, and ongoing monitoring, similar to other business-critical software.

Prasanna Arikala, CTO and Head of Products at Kore.ai, said agent adoption has accelerated and created a management challenge.

"AI agents are rapidly becoming the new software workforce inside enterprises," Arikala said. "But without centralised governance, enterprises risk losing visibility and control over how AI operates across the organisation. The Agent Management Platform introduces a new operational layer for enterprise AI, giving leaders the ability to manage AI agents with the same discipline, transparency, and accountability as any other critical business system."

Broader positioning

Raj Koneru, CEO and founder of Kore.ai, linked the launch to AI's shift from pilots into mainstream operations. He said many enterprises have moved beyond chatbots and single-purpose automation to agent-driven workflows spanning departments and systems, increasing the need for governance, monitoring, and accountability across a growing set of deployments.

"AI is quickly becoming core infrastructure for how enterprises operate," Koneru said. "But scaling AI responsibly requires more than powerful models; it requires governance, visibility, and accountability. With the Agent Management Platform, we are helping enterprises turn AI from isolated experiments into a trusted, enterprise capability that delivers real business value."

AMP is designed for multi-agent environments and heterogeneous AI ecosystems. Kore.ai says it provides a central foundation for governing AI adoption as it expands across an organisation.