Introducing our F.A.C.T AI engine

Introducing our F.A.C.T AI Engine

Why our AI doesn’t “summarise meetings”

Let’s be honest. Most AI meeting tools do one thing: They summarise conversations.

Which sounds great… until you realise something important.

In financial advice, summaries aren’t actually the goal.. We need records.

Accurate adviser file notes become part of the client record and compliance trail.

Records that capture:

• what the client said
• what the adviser explained
• what financial figures were discussed
• what decisions were made
• and what happens next

That’s not a summary.

That’s a file note.

And a proper file note is closer to structured documentation than it is to a casual recap.

So when we started building AI meeting notes for financial advisers, we quickly realised something:

If you ask a single AI model to “summarise a meeting”, it will happily produce something that sounds reasonable… but may quietly drop the most important things.

Like:

• the $1,170,000 super balance
• the $23,000 Hub24 account transfer
• the 4% pension withdrawal discussion

And if you’re relying on that summary as a client record… that’s not ideal.

So we decided to build something different.

Introducing the FACT Engine

Instead of relying on one AI step, our meeting note system runs through a structured process we call the FACT Engine. FACT stands for:

Find → Anchor → Confirm → Tell

In other words, the AI doesn’t jump straight to writing.

It works through the conversation systematically before writing anything.

Step 1 - Find the Facts

The first model analyses the data (and for the purpose of this blog, let's assume the data is a transcript) and identifies the actual facts discussed during the meeting.

This includes things like:

• financial figures
• super balances
• withdrawals and contributions
• property ownership
• strategy explanations
• client decisions
• agreed actions

Instead of producing a summary, the system creates a structured fact layer from the conversation.

Think of it as the AI highlighting the important lines in the meeting before it writes anything.

Step 2 - Anchor the Evidence

Next, those facts are anchored back to the transcript.

For example:

• a dollar amount is linked to where it appeared in the conversation
• a strategy explanation is tied to the adviser’s explanation
• a decision is connected to the moment the client confirmed it

This creates a reference index that allows the system to trace information back to the source conversation.

Which is important for one simple reason - If the AI can’t find where something was said… it probably shouldn’t be writing about it.

Step 3 - Confirm the Details

Once the facts are anchored, the system verifies them against the transcript.

In other words, the AI asks:

“Can I actually prove this was said?”

If the answer is yes, the fact is verified. If the answer is no, it doesn’t get used.

This step dramatically reduces the risk of the classic AI problem: hallucination.

(Which is a polite way of saying making things up.)

Step 4 - Tell the Story

Only once the facts are verified does the system generate the final adviser file note.

At this point the AI isn’t trying to “figure out what happened”.

It already knows.

So it can focus on doing what it does best: articulating the conversation clearly.

The final result is a structured file note written in adviser language:

• “I discussed…”
• “I explained…”
• “The client confirmed…”

Exactly the way advisers typically document meetings.

Why This Matters

Most AI meeting tools try to do everything in one step:

listen → summarise → done

The FACT Engine takes a different approach:

Find the facts

Anchor the evidence

Confirm the details

Tell the story

By separating the process into specialised stages, the system produces meeting notes that are:

• more accurate
• more complete
• easier to review
• and far more suitable for adviser records and financial planning documentation.

Which ultimately means less time writing file notes… and more time doing actual advice work.

The Bigger Picture

The FACT Engine isn’t just about meeting notes.

It’s the architecture behind how we’re building AI across our platform.

Because whether you’re generating:

• meeting notes
• strategy explanations
• client communications
• or compliance records

the same principle applies. Before AI writes anything, it should first understand the facts.

And yes… the acronym was intentional.

Find the facts.
Anchor the evidence.
Confirm the details.
Tell the story.

Or put another way:

AI that deals in FACTs.

From Documentation to Real Time-Wealth

Ultimately, the goal of the FACT Engine isn’t just better AI.

It’s more time back.

When meeting notes are automatically generated, verified and recorded directly in your CRM, advisers no longer need to spend evenings or weekends reconstructing conversations and writing file notes.

Instead, accurate documentation simply appears in the client record, ready for review, refinement and compliance. That means less administrative catch-up and more time where it actually matters.

More time to:

• spend with clients
• focus on advice strategy
• grow the practice
• or simply finish the day earlier

Because the real benefit of better systems isn’t just efficiency.

It’s Real Time-Wealth, giving advisers back the time they’d rather spend doing the work they enjoy… or even heading out for a surf before sunset.

If you're curious how the FACT Engine works in practice, the best way to understand it is to

see it generate a real adviser file note from a meeting transcript. So book a meeting now to see how we can assist.