The best analysts know how to ask the right questions of their data.
AI doesn't replace that instinct. It amplifies it.
Socratics Compose flags a 130% revenue anomaly in seconds. That used to take an experienced analyst hours to find.
Better questions. Faster answers.
The due diligence workflow hasn't fundamentally changed in 20 years.
Every data-driven industry eventually hits a moment where the messy manual foundation gets rebuilt.
Finance is at that moment right now.
There's a moment in every deal where the model stops being a draft and becomes the source of truth.
No one announces it. It just happens.
But the model didn't start with clean data. It started with whatever was available, cleaned as fast as possible.
The dangerous part: it's never escalated. Never tracked. It just silently compresses the time available for actual judgment.
The model gets built. The memo goes out. But the foundation was rushed, and no one knows exactly where.
That's what keeps me thinking about this.
Every analyst I talk to describes the same thing.
They call it "getting the data ready." Not a problem. Just part of the job.
But when you add it up: pulling line items from PDFs, reconciling periods, hunting footnotes, fixing broken exports.
That's 60-70% of the time on a deal.
Everyone's building AI that generates faster outputs.
Nobody's fixing the inputs.
Faster garbage is still garbage.
The teams that win private markets won't have the most AI tools. They'll have the cleanest data foundation.
That's what we're building at @SocraticsAI
Most people think AI will replace financial modeling.
It won’t.
It’s changing how models are built:
Clean inputs
Standardise structure
Reconcile inconsistencies
Validate outputs
The real unlock isn’t automation — it’s trust.
And that’s exactly where Socratics focuses.
As we get ready to launch https://t.co/W0D1AVNBfw, that lesson still guides us.
We give teams a clean, structured financial foundation they can understand and trust.
Excited to share more soon.
Some of my strongest memories from Wall Street aren’t from closing deals.
They’re from sitting late at night with a spreadsheet open, trying to make sense of a company’s financials before an important meeting.
(thread) 👇
That experience stayed with me.
Years later, as AI agents are becoming central to finance, I kept thinking about those nights.
The problem was never the calculation. It was the clarity of the inputs.
Alysa Liu’s 2026 Olympic gold in women’s figure skating is more than an amazing personal performance; it is an inspiring victory of freedom over tyranny.
America is more than our economy and military. Our values make the United States exceptional.
I didn't expect to come back to this problem.
After 15 years in M&A and PE, I moved to Silicon Valley to build tech companies. I assumed someone would have solved financial modeling by now.
When I returned to advising in 2023, I was shocked 👇
That's @SocraticsAI.
Institutional-grade validation for private-company financials. Infrastructure-level change for an industry that's been waiting 15 years for it.
We're honored to be named Best AI for M&A Startup at M&A Tech Connect.