The agentic close is an AI-driven approach that transforms the traditional month-end sprint into a continuous, governed process. KPMG helps companies enable this capability in a command center that coordinates specialized AI agents to monitor transactions, automate reconciliations, and provide real-time visibility over their existing finance systems.
The financial close has long been defined by pressure, complexity, and a race against time. Instead of waiting for period-end forward-thinking finance leaders can tap into artificial intelligence (AI) agents for an orchestrated process to close the books.
This is the agentic close, where AI agents continuously monitor source transactions as they occur, proactively resolve accounting implications, and offer visibility into close readiness continuously.
Ready to break the month-end cycle for good? Dive into our latest thinking on building a more resilient and proactive finance function.
Most close issues don't start during the close; they begin with the daily transactions that feed into it. Traditional tools fail under production complexity. Delays compound when activities are spread across separate systems, geographies, and manual spreadsheets.
In contrast, the agentic close involves continuously governed transactions. Rather than relying on a single tool or general-purpose AI assistant, our solution coordinates a team of specialized AI agents working together within defined risk and control boundaries.
Agentic capabilities are starting to have a material impact on the finance function. It’s no longer about automating steps; it’s about stating the goal and overseeing AI as it reasons the path to the outcome.
Dmitry Levin
Partner, Advisory, Finance Transformation, KPMG US
You’re one step away from seeing the future of the financial close.
Visibility and control
Real-time visibility into transaction quality metrics, close-readiness indicators, and emerging variance commentaries.
AI quality and trust metrics
Observe how we monitor agent performance, calibration, and reduce AI hallucinations through semantic and ontology layers.
The AI workspace
See specialized agents (accruals, reconciliations, and anomalies) coordinate and route work directly across simulated ERP systems.
Speed and quality are not competing objectives in the financial close. A continuous, exception-based approach enables finance to accelerate the close while strengthening control, consistency, and confidence in the outcome.
John Whalen
Managing Director, Advisory, Finance Transformation, KPMG US
Q: What is the “agentic close” in finance?
A: The agentic close is an AI-driven approach to the traditional financial close that shifts controllership from a reactive, period-end sprint to an “always-ready” continuous process. Instead of waiting until month-end to reconcile ledger data, a team of specialized, coordinated AI close agents continuously monitors transactions, flags accounting exceptions as they occur, and prepares draft entries.
Q: What is the close command center?
A: The close command center is a secure, centralized digital operating environment and dashboard. It acts as a control tower, orchestrating workflows across the R2R lifecycle. By layering over your existing ERPs, it connects transactions, accounting, consolidation, and controls to provide CFOs, controllers, and accountants with real-time, persona-driven visibility into close status, risk metrics, and variance insights.
Q: For whom is the close command center designed?
A: Leaders and members of finance and operations teams benefit from access to the close command center, including the chief financial officer (CFO), chief revenue officer (CRO), chief operating officer (COO), chief accounting officer (CAO, controller, senior vice president of finance, and senior vice president of global business services (GBS).
Q: How do AI close agents integrate with ERP platforms?
A: Our approach to agentic close architecture does not require a costly “rip and replace. We use a secure integration fabric with more than 200 pre-built cloud connectors to integrate directly with existing ERPs. The AI agents securely ingest data via governed API pathways, orchestrate complex workflows horizontally across these systems, and report actions back to the central command center.
Q: Is there still a human role in an automated, agentic financial close?
A: Yes. The agentic close does not replace human judgment; instead, it adopts an exception-first operating model. AI agents automate repetitive transaction matching and routine accrual calculations, surfacing only policy exceptions, variance anomalies, or complex judgments to a designated professional. Mandatory human-in-the-loop validation, strict approval thresholds, and segregation of duties are embedded directly into every agent workflow.
Discover KPMG CFO Real Insights, designed to help improve business performance across the enterprise and in your finance organization.
Ready to accelerate your close? Let’s talk about how to equip your finance team with the capacity to stop chasing yesterday’s exceptions and start driving strategic business outcomes.
