Pleo, Europe’s one of the first and most trusted strategic finance platforms, on June 11th, announces a suite of AI agents that handle routine spend management autonomously, so finance leaders can focus on decisions rather than administrative tasks.
Finance teams are under growing pressure: 45% of European companies say their current processes slow them down, and leaders describe approval workflows as so cumbersome they struggle to operate at the required speed. Meanwhile, 73% believe the future of finance teams and professionals depends on the comprehensive understanding and use of AI, yet 63% say AI skills, training and confidence are severely lagging.
Pleo’s agentic AI addresses this gap. Starting with beta testing in July, Pleo is moving toward a model where specialised agents execute routine work autonomously, surfacing only the exceptions that genuinely need human judgment.
Jeppe Rindom, CEO and Founder of Pleo, comments:
“Agentic AI gives finance leaders a clear path out of the admin and back to the work that requires their strategic judgment. We’re focused on unifying users’ full financial ecosystem-policies, budgets, vendors, approvals-while keeping data encrypted and controlled to move towards true autonomous spend: agents deciding independently, always within your guardrails. That’s when finance leaders get their profession back.”
From workflows to agents
Pleo is building specialised agents to handle financial workflows:
- Policy Agent — Currently a live agent that enforces company spend rules in real time, flags policy exceptions and routes approvals to the right person
- Pleo MCP — Next to launch feature that captures card transactions, finds receipts, writes memos, applies accounting codes and submits expenses
- AP Agent — Processes invoices from email, codes them, schedules payment and tracks approval and payment status end to end
- Treasury Agent — Monitors cash flow, tracks spend against budget and surfaces overspend risk before it happens
- Accounting Agent — Codes transactions, reconciles accounts and closes books, flagging only what needs human review
Two ways to access Pleo AI
The Pleo agents will live throughout the Pleo app and web experience. Additionally, Pleo’s MCP server acts as the infrastructure layer that lets finance teams connect their preferred AI tools – such as ChatGPT Codex, Gemini, Claude Cowork, and Copilot – so they can directly access data in Pleo. Finance teams can query spend, approve expenses, prepare bookkeeping, and run reports without switching tools.
Marija Nakevska, CPTO, Pleo adds:
“Finance software is shifting from passive dashboards to AI agents that actively work on behalf of users, safely and transparently. Over the past decade, Pleo has built one of Europe’s richest datasets from its customer base of 40,000 businesses, revealing patterns in how and where they spend, and where risks emerge. This intelligence drove Pleo’s evolution from expense management to full cash management. Now we’re channelling it into the workflows finance teams need most. They want AI working for them, delivering better control. We’re seeing real appetite for AI features, but paired with a critical question: ‘How does this use my data, and do I stay in control?’ The MCP server foundation answers both concerns. It enables advanced AI capabilities while keeping customers in control of their data.”
A key differentiator of Pleo’s agentic experience is the conversation design. Today, many AI models mimic human emotions in their responses, which can mask issues. Pleo’s agentic experience is designed for competence, not charm. This is part of Pleo’s vision to build trust at every layer.
More related news:
FPT and Shinhan Bank to Expand Collaboration in AI-Powered Finance and Startups
Pepper Introduces AI-Driven Price Management for Independent Food Distributors
