AI in Finance: Insights from the YFM portfolio

AI is everywhere right now.

Every tool, every headline, every conversation seems to come back to it. But for many CFOs and finance leaders, the question isn’t what is AI? …. it’s much simpler:

As part of our portfolio webinar series, we brought together finance leaders and AI specialists to cut through the noise and focus on what matters in practice, where AI is already being used, what’s working, and how businesses are starting to embed it.

What came through clearly is this:

AI isn’t about transformation overnight.
It’s about solving real problems, one step at a time.

One of the strongest themes from the session was a simple mindset shift.

AI shouldn’t be the starting point.

Instead, the focus should be:

• Where are the bottlenecks?
• Where are the manual processes?
• Where is time being lost?

Only then does AI become relevant.

For most businesses, the biggest opportunity isn’t building complex AI models it’s removing friction from existing workflows.

In practice, that means:

• Automating repetitive finance processes
• Improving reporting speed and accuracy
• Reducing reliance on manual spreadsheets

This is where the fastest and most tangible returns are being seen.

Across the portfolio, a clear pattern is emerging in how finance teams are adopting AI.

It typically follows a simple progression:

1. Automate manual processes

The “low-hanging fruit”

• Reporting workflows
• Data extraction
• Invoice and order processing
• Board pack preparation

These are often time-intensive, repetitive, and prone to error, making them ideal for automation.

In some cases, businesses are reducing processes from hours to minutes, freeing up significant capacity.

2. Build reliable, automated reporting

From spreadsheets → to real-time visibility

Many finance teams still rely heavily on:

• Excel-based reporting
• Manual data consolidation
• Static dashboards

The shift here is towards:

• Automated reporting suites
• Connected data across systems (finance + CRM)
• Faster, more reliable access to numbers

The goal is simple:

To trust the numbers and access them instantly.

3. Unlock insight from data

From reporting → to decision-making

Once the foundations are in place, the next step is using data more proactively.

This includes:

• Customer lifetime value analysis
• Cost of acquisition insights
• Identifying profitable segments
• Spotting trends that aren’t immediately visible

Interestingly, this is often where assumptions get challenged, with data disproving long-held “gut feel” views about what drives performance.

One of the more recent shifts is how leaders interact with data.

Traditionally:

• Fixed dashboards
• Monthly reporting cycles
• Static board packs

Now:

• Leaders can ask questions directly of their data
• Generate answers and visualisations in real time
• Interrogate performance during meetings

This moves finance from reporting what happened to exploring what’s happening now.

Rather than large-scale transformation programmes, most businesses are taking a pragmatic approach.

Some examples discussed in the session included:

• Automating contract and proposal creation (with human review)
• Reducing operational workload by up to 50% in some teams
• Streamlining internal processes to remove bottlenecks
• Using AI to support, not replace, decision-making

A consistent theme:

Or as one contributor put it:

Aim for 80% automation, with humans focused on the 20% that adds real value.

The biggest barrier isn’t technology.

It’s time.

For most CFOs and leadership teams, the challenge is:

• Competing priorities
• Day-to-day operational demands
• Knowing where to start

Which is why the most effective approach we’re seeing is:

Start small. Prioritise. Build momentum.

Where to start: a practical framework

For teams looking to move forward, a simple approach is emerging:

1. Define your principles

• How will you use AI?
• What are your boundaries (e.g. data, ESG, compliance)?

2. Set a clear toolset

• Approved platforms
• Avoiding fragmented or “shadow” usage
• Ensuring data security

3. Focus on real use cases

Prioritise based on:

• Effort vs impact
• Time saved
• Value created

4. Upskill continuously

AI capability isn’t a one-off training exercise.

It requires:

• Ongoing learning
• Practical use
• Building confidence across teams

5. Build a roadmap

Not everything at once.

Just:

• The next few priorities
• Delivered consistently

For PE-backed businesses, this ultimately ties back to one thing:

Value creation.

AI can support this in three key ways:

• Increasing profit → through efficiency and automation
• Improving decision-making → through better insight
• Enhancing valuation → through innovation and scalability

But importantly:

AI is not the strategy.

It’s an enabler.

AI can feel complex, fast-moving and sometimes overwhelming.

But the takeaway from this session was refreshingly simple:

You don’t need to do everything.

You just need to start.

Focus on:

• One process
• One problem
• One improvement

Because in most cases, the biggest gains aren’t coming from cutting-edge AI.

They’re coming from doing the basics better, faster, and more consistently.