The frontier lab building Large Tabular Models for enterprise prediction
San Francisco
- Tweets 66
- Following 3
- Followers 309
- Likes 26
Pinned Tweet
What if we could more accurately predict what comes next? That's the question that started Fundamental.
A couple of years later, it’s changing what AI can do.
3
164
In F1, you get very few laps to understand whether you’ve got the car right.
So we put NEXUS to work predicting front aero balance before the car even left the garage, using track, weather and car setup data.
𝟭𝟬.𝟯% 𝗯𝗲𝘁𝘁𝗲𝗿 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝘃𝗲 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲.
That’s a huge result in a sport measured in thousandths of a second - giving engineers a better understanding of the car before they spend a single lap learning it on track 👇
fundamental.tech/case-studie…
1
53
San Francisco. Barcelona. Singapore. And now Tel Aviv.
We’re very excited to announce we've opened our new office in Tel Aviv.
We started Fundamental to build a new kind of foundation model - one built for the tabular data that powers the world’s businesses. Doing that means bringing together exceptional people wherever they are.
Four cities. Three continents. One team building something entirely new.
Join us 👉 fundamental.tech/careers
1
2
96
Fundamental retweeted
Introducing VAST DataEnclave. Built on NVIDIA Confidential Computing, this new runtime in the VAST AI Operating System enables leading models to securely run against sensitive enterprise data while protecting customer and model IP.
4
3
10
3,349
Seeing is believing. Two real-time demos, powered by NEXUS, on display this week at Booth #214 at @Dreamforce.
🛍️ Retail IQ: Demand forecasting, churn, and pricing elasticity staged directly by Agentforce for human approval.
📉 Deal Slip Intelligence: Real-time slip risk scores, driver rankings, and what-if simulations embedded right in Salesforce.
2
135
Fundamental retweeted
If you’re in SF for Dreamforce this week, you might spot a few Large Tabular Models (LTM) around town 👀🇺🇸
@LTMpredict @Dreamforce
1
1
5
342
Day 1 of the Buzz-O-Meter is in: "Agentic" was said 259 times on stage 🎯
Today we're tracking "AI", a word NEXUS predicted by analyzing 900k+ transcript words and 46k headlines.
Stop by our Howard St. booth, predict today's count, and win merch 👇
buzzometer.fundamental.tech/…
105
This week at @Dreamforce we’re making the case that AI has been working with half a brain.
LLMs changed what’s possible with language. But the decisions that run the enterprise - what customers will do, what demand will look like, what will fail, what happens next - live in tables.
That needs a different kind of AI.
We’ll be all over Dreamforce this week. Come find us:
📍 Booth #214: See NEXUS in action
🎙️ Tuesday, 3:30 PM: CEO @fraenkelj on stage at the SMB Keynote
🧠 Howard St.: Put your prediction skills to the test with our Buzz-O-Meter
1
190
LLMs were never built to predict what happens next from enterprise data. Large Tabular Models were.
And we think they’re about to become as fundamental to business as LLMs have become to language.
fundamental.tech/news/the-se…
1
103
Fundamental retweeted
Recently, I had a great chat with my friend and former colleague @GarneloMarta, CSO at Fundamental (@LTMpredict).
We talked about diffusion models (duh 🤭), music generation, typicality, blogging, @kaggle and more (also restricted Boltzmann machines 🫣)
fundamental.tech/news/unders…
6
18
191
14,432
"Tabular data has never been front and center in the industry, but the reality is that this is what drives businesses."
That's Léo Dreyfus-Schmidt (@dsleo), who spent eight years running Dataiku's research lab and watched the field move through images, then reinforcement learning, then language models. He joined Alex Gerbeaux, Head of Applied AI at Fundamental, on the latest installment of Ground Truth to talk about where tabular data sits now that the rest of the stack has moved.
fundamental.tech/news/leo-dr…
2
281
Fundamental retweeted
somewhere right now an exploration and production leadership team is staring at a dashboard, deciding how much capital to commit to a field based on subsurface models that leave half the picture in the dark. they are guessing. and they are good at it. but they are guessing. multiply that one guess across every operator, every basin, every hour, and you've found one of the largest hidden costs in the world economy.
1
2
4
930
Oil & gas might be the most underrated AI opportunity on the planet.
It’s an industry that's spent over 100 years trying to predict what happens next.
How much a field will produce.
How quickly a well will decline.
When equipment will fail.
Where risk is building.
And over the decades it’s accumulated a massive amount of data on how all of it actually played out.
1
2
3
278
It's an extraordinary position to be in right now. We've already seen one major operator report over $500M in value from AI in a single year. With this kind of predictive capability applied across thousands more decisions like production, reliability, maintenance, safety, planning, it'll be incredible.
1
156
There are few industries better positioned for what comes next.
More on why, ahead of @ONS_Stavanger ↓
fundamental.tech/news/oil-ga…
1
57
Our Founder & CEO @fraenkelj joined @Stanford this weekend for a panel on The Future of AI Beyond the Chatbot Era
Alongside Igor Babuschkin @ibab, Ricardo Baeza-Yates @PolarBearby, and Lu Zhang @luzhangvc, the discussion explored where AI goes next, from specialized systems and autonomous agents to trust, ownership, and what comes after today's generation of chatbots
1
9
318