Resolution Agents are AI-powered accounting automations that automatically generate corrective journal entries to fix discrepancies found during reconciliation. They complete the automation loop by turning detected issues into auditable, structured fixes, helping finance teams close faster and with fewer errors.
Finance teams have invested heavily in tools that detect problems. Matching engines flag discrepancies, anomaly detection surfaces misclassifications, and reconciliation platforms highlight what does not tie out. But here is the uncomfortable truth: finding the issue was never the hardest part. Fixing it is.
Every month, controllers and senior accountants sit down with a list of identified discrepancies and begin the slow, repetitive work of crafting corrective journal entries. They reverse mismatched transactions, post clearing account adjustments, record timing differences, and document every step for audit. It is precise, high-stakes work that demands accounting expertise, yet most of it follows patterns the team has executed dozens of times before.
This is the reconciliation resolution gap, and it remains one of the biggest bottlenecks in the month-end close. Resolution Agents were built to close that gap. Within Nominal's Agentic Performance Management (APM) model, where autonomous systems execute complete accounting workflows, they handle the critical final step: automatically generating corrective journal entries for discrepancies identified during reconciliation, completing the loop that turns detection into action.
What Are Resolution Agents?
Within Nominal's agent ecosystem, these specialized systems automatically generate corrective journal entries to resolve discrepancies identified during reconciliation. When a Matching Agent flags an unmatched transaction or Transaction Patrol detects an anomaly, the finance team investigates and determines the right course of action.
The fix is then executed, producing a properly structured, auditable journal entry with correct debits, credits, account mappings, and dimensional accuracy.
Think of them as the final piece in a complete reconciliation workflow. Detection tools find the problem. The finance team decides how to fix it. This automated execution eliminates manual entry creation while keeping human judgment at the center of every accounting decision.
The Reconciliation Resolution Gap
Detection without resolution is an incomplete workflow. When reconciliation tools identify 40 discrepancies but every single fix requires manual journal entry creation, finance teams are left with the most time-consuming part of the process untouched. The issues are visible, but the resolution still depends entirely on human execution.
This creates a set of compounding challenges:
- High volume, repetitive corrections consume hours of skilled labor. An inventory clearing reconciliation alone might require hundreds of individual adjustments, all following identical logic.
- Manual execution introduces errors because copying amounts, reversing signs, and mapping accounts under deadline pressure lead to mistakes.
- Documentation becomes inconsistent when different team members record resolutions in different formats, creating gaps that auditors will question.
- Time pressure at month-end means all of this work concentrates exactly when teams face maximum constraints and minimum margin for error.
The resolution gap is not a minor inefficiency. It is a structural bottleneck that delays close cycles, increases audit risk, and keeps senior accountants trapped in mechanical work instead of strategic analysis.
You might also like: AI in Audit: Automating Reconciliations and Financial Reporting
What It Looks Like Without Automation
Without automation, a controller manually resolves a discrepant intercompany match, like reversing Side A and posting to an FX variance account, by performing multiple steps in the ERP, taking 10 to 15 minutes per resolution.
Resolving 40 discrepancies requires 6 to 10 hours of repetitive work.
With a Resolution Agent, the controller determines the resolution in Nominal's interface, selects the pre-configured agent (e.g., "Reverse Side A with FX Variance"), reviews the draft entry, and approves it. This takes 2 to 3 minutes per resolution.
The same 40 discrepancies now take only 1 to 2 hours, achieving an 80% reduction in resolution time with perfect accuracy and a complete audit trail.
Why APM Outperforms Traditional Automation
This complete loop is what separates APM from legacy tools. Previous approaches made processes slightly faster but failed to change how the work actually gets done.
- RPA (Robotic Process Automation) often breaks when data formats shift or UIs change.
- AI Assistants can suggest answers, but typically cannot execute entries in the GL.
- Workflow Tools simply rearrange the steps without eliminating the manual accounting work.
How Nominal's Resolution Agents Work
Agentic Performance Management introduces autonomous execution into finance. Instead of tools that assist accountants or automate singular tasks, APM utilizes programs that complete entire business processes from start to finish. Resolution Agents exemplify this by managing the corrective entry procedure entirely within Nominal's platform.
This is possible because of Nominal's three-pillar architecture: a native General Ledger for transaction-level data access, comprehensive Task Management for human oversight, and embedded AI Agents for automated execution.
Together, these components give the resolution solution the full accounting context they need to generate accurate, structured outputs.
The workflow follows five clear steps, each designed to keep accounting judgment with the finance team while removing mechanical execution.
Step 1: Issue Detected
A Matching or Transaction Patrol Agent identifies a discrepancy, whether that is an unmatched intercompany transaction, a misclassification, or a balance that does not tie out, and flags it for the finance team.
Step 2: User Investigation
The team opens a task to investigate the root cause. They collaborate, analyze the underlying transactions, and determine the appropriate resolution approach. This is where human judgment drives the process.
Step 3: Resolution Agent Selected
Teams can configure multiple agents for different resolution patterns. A company might have separate systems for inventory variance posting, FX reconciliation adjustments, timing difference clearance, and intercompany elimination corrections. The controller selects the one that matches the scenario.
Step 4: Resolution Agent Executed
The entity receives full context about the match or issue being resolved, including all transaction lines, amounts, accounts, entities, and dimensions. It interprets the configured logic and generates a properly structured journal entry with correct debit and credit inversions, balancing amounts posted to the right accounts, and full dimensional accuracy.
Step 5: User Review
The generated journal entry appears as a draft within the same task. The controller reviews the proposed resolution, can override or modify if needed, and signs off. Upon approval, the entry posts to the ERP, the task updates to resolved status, and the complete audit trail is permanently documented.
This end-to-end flow reflects the core APM principle: automation handles execution while humans govern outcomes.
Natural Language Configuration
Instead of writing code or configuring rigid rule engines, finance professionals express resolution instructions in plain language. For example: "Reverse all matched lines and post the balancing amount to account 5100" or "For timing differences under $1,000, post to account 4200; for amounts over $1,000, post to account 4300."
Controllers and accounting managers can create, adjust, and refine system behavior without engineering support. When business logic evolves or new scenarios emerge, the team updates instructions in the same language they would use to explain the process to a colleague.
Context Aware Execution and Validation
Because they are embedded within Nominal's platform, Resolution Agents access the full context of every issue they resolve. They work directly on accounting data structures, understanding accounts, entities, dimensions, consolidation hierarchies, and historical patterns through the native General Ledger.
Before any recommendation reaches a controller, internal validation critics review the output. If a generated entry has debits that do not equal credits or violates accounting structure requirements, the critic blocks it. This ensures that every draft entry surfaced for review meets accounting integrity standards before a human ever sees it.
Why This Matters for Finance Teams
The impact of Resolution Agents extends well beyond time savings, though the time savings alone are significant. Teams that previously spent days creating corrective entries now complete the same work in hours.
Faster close cycles
Automated resolution removes one of the most persistent bottlenecks from the month-end close. When corrections that took 6 to 10 hours now take 1 to 2 hours, close timelines compress meaningfully.
Improved accuracy
Resolution logic applies identically across all similar discrepancies. There are no transcription errors, no variations between team members, and no fatigue-driven mistakes during high-pressure close periods.
Stronger audit posture
Every resolution carries a complete audit trail: what issue was detected, how it was investigated, which take-out was invoked, what entry was generated, and who approved it. Auditors get transparent documentation without the team spending extra time creating it.
Institutional knowledge capture
Resolution patterns encoded in processes do not disappear when team members leave or processes change. New staff can review process configurations to understand exactly how the team resolves different types of discrepancies.
Scalability without added headcount
The same agent logic that handles 10 monthly corrections handles 100 or 1,000 without additional configuration. As organizations grow in complexity, resolution capacity scales with them. This is the APM advantage: finance operations that expand without proportional staffing increases.
Recommended read: How to Scale Finance Operations Without Adding Headcount: Data-Driven Insights from 50 Million Transactions
The reconciliation resolution gap has been one of the most persistent and costly blockers in finance. While detection tools surfaced issues faster, resolution remained manual and error-prone. Resolution Agents change that by automating corrective entries while keeping finance teams in control. The result is a faster, cleaner, and more strategic close.
See how Resolution Agents close the loop between insight and execution. Book a demo to experience end-to-end reconciliation automation with Nominal.
