AI in the finance tech stack: CFOs strategies from Money 20/20

At Money20/20, we sat down with three forward-thinking finance innovators to explore how AI is already making an impact in real-world finance stacks today — and to uncover where this fast-moving technology is heading next. Here are our biggest takeaways:

What is actually embedded in finance tools?

AI is already embedded in the daily lives of modern finance teams — not as a standalone solution but woven into existing tools.

Andrew Rudchuk, Head of Finance Transformation at Starship Technologies, a fast-growing delivery robot business, shared how platforms like Confluence, Jira, Pigment, and NetSuite are now enhanced with embedded AI features. From document generation to forecasting tweaks, AI is helping finance leaders speed up insights and streamline processes.

Meanwhile, Ilya Shabrov from European micro-mobility company, Dott, took it a step further. Ilya explained how his team has been working with Payhawk to use OCR and predictive models since day one. Their early AI tools learned from user behaviour to categorise invoices and extract VAT data. And now, they’ll scale that vision with conversational Payhawk AI agents that understand travel policies and guide employees through company-approved expense flows.

Building autonomous finance agents

Where things got even more interesting was within the shift from "assistive AI" to autonomous agents.

At Starship, Andrew described building a "digital twin" of the finance function. A complete virtual model of processes, data, people, and costs. This twin is analysed using LLMs to detect bottlenecks and generate strategic recommendations, like a second brain for your back office.

Meanwhile, Konstantin Dzhengozov (CFO and co-founder at Payhawk) explained how procurement typically touches multiple teams, such as legal, finance, and compliance. And with AI, businesses can orchestrate the entire workflow: Extract relevant clauses from a contract, notify legal and data privacy stakeholders, raise a PO, trigger accruals, and handle sign-offs.

This way, AI isn’t just doing tasks. It’s managing coordination across departments.

Adapting to local compliance

Starship's Andrew also introduced a vision for AI-native systems that can dynamically adapt to jurisdiction-specific requirements. Instead of hardcoding tax logic into tools, you prompt an agent, saying: "When a PO is approved, create a PDF, email the supplier, match incoming bills, and handle the payment."

The agent takes care of the logic and even adjusts for local tax regulations automatically.

This kind of localisation-as-a-service removes a huge operational burden from finance teams and eliminates the need for painful ERP customisations.

2026: The risk of standing still

The panel closed on a powerful message. Standing still is no longer neutral... It’s a risk.

Andrew from Starship shared:

In 2026, finance teams that fail to invest in AI risk becoming operationally stagnant.

All three panel lists agreed. The highest-performing teams of the future will be the ones that experiment early. The shift doesn’t have to come top-down. In fact, some of the most impactful use cases come from the ground up. Controllers and analysts are testing AI prompts to streamline their own workflows. Andrew continued:

The best use of AI will come from the knowledge workers who understand the data and the challenges.

What CFOs should do today

AI in finance offers many growing benefits. But still, it's not about completely replacing your current ERP system immediately. It’s about making AI a team sport across finance.

Here’s where to start:

• Audit your stack. What tools already offer embedded AI? What manual work could be offloaded?
• Empower AI champions. Give your team space to experiment, play, and build
• Start with repeatable workflows. Automate things like PO approvals, expense categorisation, and policy enforcement. (Because, boring AI is better AI for finance)
• Think long-term: AI agents aren’t just helpers, they’re co-workers in the making

Want to see how Payhawk is putting these ideas into practice? See how our AI agents are helping finance teams reclaim time, reduce admin, and stay in control.