Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?

Hi {{first_name|everyone}},

In this issue:

🤿 DEEP DIVE

The model matters, but the data might matter just as much.

A smarter AI model is nice. But if you’re a banker and half your day is spent finding the right numbers, checking where they came from, and getting everything into the right spreadsheet or deck, intelligence alone doesn’t really solve the problem.

That seems to be the thinking behind ChatGPT for Financial Services.

OpenAI has built a version of ChatGPT Work specifically around financial services, shaped through design partnerships with Morgan Stanley and Evercore. It includes GPT-6 Astra, but honestly, the more revealing part is everything OpenAI has put around the model.

They’re trying to remove the data wrangling

ChatGPT for Financial Services includes premium datasets from Daloopa, PitchBook, LSEG News, and Crunchbase. That covers things like earnings transcripts, financial statements, company fundamentals, and private company data.

And these aren’t just connectors sitting off to the side. OpenAI says the data is indexed and hosted on its own infrastructure, which lets it improve retrieval and latency while giving users granular citations back to the source.

The example in the announcement is a good one. If a banker is normalizing a P&L, they could inspect the reconciliation and notes behind adjusted EBITDA, see which costs were excluded, and then decide how to use that figure in a valuation.

That’s a pretty good illustration of what this product is actually trying to do. Not just produce an answer. Let someone follow the number back through the evidence behind it.

For firms that already pay for financial data, OpenAI is also working with S&P Capital IQ, LSEG, MSCI, Dow Jones Factiva, and Moody’s on shared sign-in and entitlement integrations. The idea is that ChatGPT recognizes what data a user is already allowed to access.

Then it has to make the actual work

This is where GPT-6 Astra comes in.

OpenAI says the model is built around three capabilities relevant here. Finding information, reasoning over financial data, and creating artifacts. It can work through figures, tables, and supporting notes, perform analysis, and turn the result into documents, spreadsheets, and slides.

Firms can even publish their own Excel, Word, and PowerPoint templates through an admin page, so the output can become a valuation model, research note, or pitchbook in the firm’s existing format and style.

So the bigger idea here isn’t really “ChatGPT for bankers.”

It’s getting the data, analysis, evidence, and final deliverable into one workflow. That seems to be the problem OpenAI and its design partners have decided is worth solving first.

The quick take

🤝 FROM OUR FRIENDS

The need for a deck rarely starts in a presentation tool — it starts in a sales call, a CRM update, or a quick ask to your AI assistant.

With Pitch’s new MCP, that’s now where it can end too. Connect your Pitch workspace to Claude once, and create, find, and share decks through plain conversation.

Ask for a proposal built from your call notes and CRM record, and get a shareable link back in the same chat. Prefer predictability? Generate from templates with variables (for customer names, dates, and pain points) to get output that matches your template, every time.

For teams automating at scale, Pitch API brings the same power to Zapier, your CRM’s workflow builder, or your own scripts — triggering personalized decks the moment a deal changes stage.

🔥 WHAT’S THE BUZZ

What people are building, sharing, and talking about

I keep coming back to the fact that this isn’t really an outsider warning about AI, it’s coming from someone whose job is to work on the alignment problem itself. And I do wonder what it means when people inside the labs are willing to say, pretty plainly, that they still don’t know whether the safety work will hold up if the systems get much smarter than us.

⚒ TOOL SNAPSHOTS

Futuristic tools within AI, no-code, and productivity

⚡ QUICK HITS

The latest in AI, tech, and productivity worth knowing

👀 ON OUR RADAR

One more thing we thought you'd like

DNSMOS gives you a score, not whether that data fits your model. Treat it as pass/fail and you'll train on audio that looks clean but hurts performance, while tossing good data for no reason. Voices' CTO DJ Jalali just published a white paper with the four-step framework the team uses to set internal thresholds instead.

Download the White Paper

🎁REFERRAL REWARDS

If TIW has been useful, invite one friend or coworker who’d actually use AI/automation at work.

Your referral link: {{rp_refer_url}}

Tip: send it to a friend who’s AI-curious, a coworker who wants to move faster, or anyone trying to automate repetitive tasks.

🤖 AI MADE THIS

Interesting and inspiring creations made with AI

ℹ️ ABOUT US

The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.

We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.

Our mission is to empower individuals, boost their productivity, and future-proof their careers.

We read all your comments - please provide your feedback!