---
id: ins_albert-malikov-pmm-as-craft-not-title
title: 'Agentic AI in finance has to work inside accounting processes, not around them'
operator: Albert Malikov
operator_role: CEO @ Stacks | AI for Enterprise Accounting | Breaking Down AI in Finance
source_url: https://stacks.ai/resources/introducing-agentic-ai-the-next-leap-for-finance
source_type: thread
source_title: 'Introducing Agentic AI: The Next Leap for Finance'
source_date: 2026-04-10
captured_date: 2026-05-04
domain: [pmm]
lifecycle: []
maturity: applied
artifact_class: framework
score: { originality: 3, specificity: 3, evidence: 2, transferability: 3, source: 3 }
tier: B
related: []
raw_ref: raw/linkedin/reactions/linkedin-reactions-2026-04-10.md
---

# Agentic AI in finance has to work inside accounting processes, not around them

## Claim

Most AI tools for finance ignore the fundamentals of accounting, controls, evidence, and accuracy, and try to bolt intelligence on from the outside. Stacks built Agentic AI that lives inside the accounting workflow itself, automating up to 90% of reconciliation work and 99% of manual journal effort while preserving complete audit traceability for every decision.

## Mechanism

Finance refuses to adopt tools that break controls. By embedding agents inside reconciliation, journal-entry, and variance workflows, not as a parallel layer, every action carries the audit trail finance teams already need. That moves the close from reactive scramble to one-click without forcing a process rewrite.

## Conditions

Holds where buyer evaluation criteria include auditability and where AI can be embedded into existing workflow software. Fails where AI is sold as a standalone copilot detached from the system of record.