It's been gratifying to follow the arc of @oneThomasli, Daniel and @JieHuang5’s journey since the beginning!
They are building the most trusted and reliable data infrastructure behind AI-Driven Finance. Comprehensive coverage of over 5,500 companies globally and used in production by analysts & agents in over 100 financial institutions!
Congrats @Daloopa team on $47M Series C! The journey continues. 🚀
@NexusVP@priyasaiprasad@ed_roman@salil@VivekGaripalli@zachweinberg@athomasq #AI
Had a blast with @detroitcoder! Went we very deep into how we are thinking about AI today in finance. Covering everything from why agents should be free and open, to orchestrating skills for specific investing outcomes.
In the Season 2 Finale of The Hedgineer Podcast, host Michael Watson sits down with Thomas Li, CEO of Daloopa, to discuss the shift from basic data extraction to complex, autonomous analytical workflows.
If you are a Wall St analyst and you did work for Mercor on the side, know that all that info might now be exposed as part of the Mercor security incident
Good read.
Important callout - AI sold to tinker-ers have exploded, but enterprise adoption takes longer. Security review, procurement, compliance etc. are real processes that are taken seriously. The largest opportunity in AI ultimately will come with enterprise deployment, and we are months away from seeing the first of many large scale agentic deployments in finance.
https://t.co/Pht1G24OGE
If you find yourself asking: can models do [x] in the future, you should read this. Pretty much lays out the reward system / interact conflicting priorities of model capabilities. This is the framework to understand what gets killed and what is valuable tomorrow.
An “agent” is not a real thing… there are skills (a set of written instructions that can be reused, kinda like the skill to ride a bike) and there are folder structures (where you keep your skill files so you can use a few skills at once). The use of skills can create an agentic experience (autonomously gets you to outcome right away). If you have a skill to build a model, and use it to build 10 models, you didn’t create 10 agents, you used a skill 10 times. Just like how if an analyst builds 10 models, you didn’t magically hire 10 people.
ryan and team worked extremely hard to make gpt-5.4 great for finance
it's much improved for financial modeling and analysis, integrates directly into excel, and connects to factiva, daloopa, s&p global, and many more
it does feel like a codex moment is coming here
Saw the unfortunate news about the layoffs at MS today. We are offering free Daloopa accounts to anyone affected to help with recruiting. You can get access to all our models / use our MCP in OpenAI or Claude for pitch and model creation etc. No questions asked.
Here is a form to fill if you need access and we will send over credentials shortly. Share with your friends if they need it.
https://t.co/4igQ46MH1a
@mstockton All skills designed to work without a daloopa license actually - it will be much more powerful and have no hallucination if you are already a customer.
Opening-sourcing another skill for finance today: this one will run a deep analysis of supply chains. it will find out all suppliers, inventory levels, gross margin risks etc for major suppliers for any company. Then it will run a stress test.