Alternative Data Weekly #254
Theme: synthetic data to systems of action
Announcement:
This is an article published Sept 1st by Rebecca Natale of Waters Technology about my former employer. Fintech powering LSEG’s AI Alerts dissolves.
As mentioned in last Friday’s ADW, as a result of this dissolution, there are 20+ high quality people available spanning engineering, marketing, project management, and GTM.
I greatly appreciate this community of people. There are a lot of good things happening in the world of data & AI. The future is bright!
Special thanks to our sponsor:
Contact Michiel Maaskant to learn more.
QUOTES
“I am still hesitant to trust AI-generated output. Why? Because financial decisions cannot be taken lightly. As a quant, I want to double or triple-check everything before I feel comfortable presenting the results. Most of the AI chatbots seem to produce “results,” but I have no way of knowing if they are correct.” - Cindy Lin as published in AI in Finance: The Next Wave of Financial Services.
News
Pods
Charts
Final Thoughts (ch-ch-ch-changes)
#1 - Adam Zewe published notes from an interview with Kalyan Veeramachaneni (Datacebo): 3 Questions: The pros and cons of synthetic data in AI. September 2025.
My Take: I’ve recently gained a greater appreciation of the value of synthetic data. What has changed recently (due to advances in AI) is the ability to create very realistic synthetic data.
There are big benefits from a governance & compliance perspective, but there is also the ability to fine tune the specific data which you want to query or test. This can reduce the cost and increase the speed at which decisions are made.
#2 – Brad Schneider published LLMs Are the New Databases (and ChatGPT Is Just the First Application). August 2025.
My Take: I’ve been thinking about this idea that systems of record, like CRM databases, are now going morph into systems of action (see Software 3.0 article I’ve highlighted a couple of times).
Much like how databases moved into the background and people only engage with application on top of databases (like a CRM), in the future “AI” will sit in the background and everything we do will be entered into the system as a fact that further informs the system what we are doing so the LLM can predict the next best action.
“Most people today don’t enter things directly into any kind of a database. They’re entering things into a highly customized application that at its core is a wrapper on a database.”
#3 – Cindy Lin published AI in Finance: The Next Wave of Financial Services. August 2025.
My Take: This is a good review of various financial AI products. I agree with the trust issue (see above quote of the week). What are the levers needed to demonstrate that the output of these AI products is trustworthy? For many of the high leverage decisions being made, even 99% accurate is not good enough.
What else I am reading:
Nate B. Jones published The MCP Implementation Guide. September 2025.
Data Action Mentor published Building your data team like a startup. September 2025.
David Wignall published Hedge Funds’ New Secret Weapon. August 2025.
Aaron Adolphson published Yodlee Launches New Era with STG Partnership. September 2025.
Samir Sharma published Classic Data Mistakes. August 2025.
Annelise Levy & Aruni Soni published Anthropic Settles Major AI Copyright Suit Brought by Authors. August 2025. Hugh O’Connor published related opinion: provenance as a competitive edge
Sven Balnojan published August 2025 wrap up & news. August 2025.
World of DaaS Roundtable Recap: Web Scraping & Data Collection Challenges. August 2025.
Source: Michael Watson of Hedgineer published From Viral Reddit Tool to Enterprise AI: The OpenBB Story with Didier Lopes | S2E2. August 2025.
Related post from Didier Lopes: From data silos to AI agents: Financial data evolution. August 2025.
My Take: Long interview (90 minutes) but very interesting. The story of OpenBB demonstrates the power of creating in public. Didier’s work with OpenBB on Reddit, specifically his work contextualizing data, drew the attention of Michael Watson (while Michael was with Citadel). As a result, Michael encouraged the progress & followed along with OpenBB’s development as an interested observer over the past few years.
Of most interest to me was their conversation the how economics of domain knowledge are changing (~minute 55). The biggest impact of “AI” is going to be taking the high performing person & 10x-ing what they know & do. The “acqui-hiring” of very talented AI-first organizations is a result (seeing this in some cases).
Highlights (92-minute run time)
Minute 02:30 – intro conversation.
Minute 04:00 – Didier’s background and description of what he has built.
Minute 08:30 – turning 4,000 Github stars into a company.
Minute 12:45 – open platform while also having a moat around structure of data.
Minute 16:00 – solving taxonomy/classification problem.
Minute 18:00 – UI discussion/evolution.
Minute 23:00 – customizable workspace & workflows.
Minute 26:00 – MCP data provider vs consumer; what do the terminal providers do?
Minute 35:30 – consumption-based pricing … Carbon Arc cited.
Minute 40:00 – 10 minutes of a fairly technical conversation about real world impact.
Minute 53:30 – Cybersyn discussion.
Minute 56:00 – Claude code sub-agents; cascading of tasks (most interesting part of pod).
Minute 60:00 – power of acquihires; talent retention (“sucks for the org, great for the person”).
Minute 65:00 – garden leave and compensation discussion.
Minute 70:00 – “internal expert networks”
Minute 75:00 – how to sell data and justify the costs?
Source: Daria Cupareanu published I Don’t Want AI to Replace My Mind. Here’s My Simple System + 17 Prompts. September 2025.
“But to understand exactly how this shift is happening, we need to look at what AI actually does to the cognitive work itself.”
“Here is a simple three-part framework that covers the what to do before, during, and after an AI conversation to ensure we're using it as a true thinking partner.”
Below is a repeat of my message from last week. I had some fantastic conversations over the past few days.
This is a wonderful community of people. Thank you.
Please reach out if you have ideas for me or the ModuleQ team.
This week, my employer, ModuleQ, announced they are winding down operations & have let the team go.
I want to acknowledge my incredible teammates, a talented group across engineering, product, project management, marketing, and GTM, who are now open to new opportunities. Any team would be lucky to have them.
So what’s next for me personally? That’s the exciting part.
I’m exploring new opportunities and leaning into what I do best: building connections, cultivating relationships, showing up consistently, and bridging the gap between technology & business.
The future of data and AI is wide open. The space is only beginning to unfold, and I couldn’t be more optimistic about the years ahead.
If you see an opportunity where I can add value, or if you simply want to swap ideas about where data and AI are headed, I’d love to connect.
John (LinkedIn)
+1 216-577-3618