AI Due Diligence: AnalystAI

Hey, it's Matt. Welcome to AI Street Markets edition, where I highlight AI investing tools. Was this email forwarded to you? Sign up below:

BACKGROUND

If you’re new here, check out the last few editions on using AI for investment analysis, creating customized news feeds and tracking earnings call mentions:

• Tracking Big Tech's AI Spending

• DeepSeek on Earnings Calls: From Zero to 100+ Mentions

• Demo: AI Agents in Market Analysis

• ChatGPT's Task Scheduling: Automated Market Analysis

• Effective Prompts For Investment Research

As a reminder, this is not investment advice, but for research purposes.

DELIVERY

You can customize how often you receive AI Street:

• Thursdays: AI news, analysis and expert interviews

• Sundays: Demos of AI investor platforms

• Both

Update your preferences here.

PRIVATE EQUITY

I talk a lot about AI and unstructured data around here. I'm always trying to think of better ways to say “unstructured data” because it just sounds so boring. But despite the name, it’s important to understand what this means and how it compares to structured datasets.

Tools like ChatGPT let you basically take any scanned document and organize its data at scale. Before, data traditionally needed to be in an Excel-type format: columns, rows, cells, etc.

The SEC requires certain filings to be in structured formats, so investors and analysts can quickly download the latest information into their models.

But not all filings are in this format, such as earnings presentations, and many jurisdictions that don’t require these formats, so analysts often spend a lot of time wrangling information from disparate documents.

Today’s edition features a demo from Ata Onat, founder of AnalystAI. His company started out as a side project, while Onat was working in private equity at EMP Belstar SuperFreeze in Southeast Asia, focused on commercial real estate.

Ata began tinkering with AI, building internal tools that helped his fund analyze targets more quickly. Once they started using his tools to close deals, he realized there was likely a broader market for the technology.

In the video below, which is ~7 minutes long, he walks through how he uses AnalystAI to create a report on Keppel, a Singapore-based global asset manager, by extracting details from images and text across multiple PDFs.

AnalystAI achieves high accuracy by using specialized vision AI models that attribute each data point to its specific document and page, according to Onat.

The system processes documents in about 5 minutes versus 2-3 days for traditional analysis, according to the company's clients. Current users include a global top-30 bank's investment banking division and funds across the U.S., UK and Singapore.

If you’re interested in chatting with Ata, please reach out at: ataonat [at] analystai.ai