Predicting the world
through numbers.

A foundation model that predicts across every industry — in a single inference pass.

Why numbers

Numbers are the language of the world.

Language lives in words; industry lives in numbers. What to price, who to treat, which unit will fail, where demand moves next — every decision that moves an industry is a question buried in a table.

What it does

One foundation model.
Every table. No pipelines.

One model, pre-trained once, reads any table and fills in the value you need — with no task-specific training and no pipeline to maintain. The workflow that once took months collapses into a single inference.

Numerical reasoning.

It sees the patterns of numbers and fills in the missing values. It answers predictive questions that LLMs cannot.

One model. Every industry.

A single foundation model that predicts across commerce, healthcare, finance, manufacturing, and defense.

Seconds, not months.

Typical ML projects take weeks to months. It takes seconds — from raw data to predictions.

Use cases

Different questions. One answer.

Across industries the questions look different, but the shape is always the same — a value missing from a table, waiting to be filled. Here is that one model, put to work.

CommerceWhat discount rate maximizes profit next month — and where does it stop paying off?
PriceDiscountProfit
$590%$128K
$5910%$151K
$5920%$139K
RetailHow many units will each SKU sell next month, given price and promotion?
SKUPricePromoUnits
SKU-19$29No4,120
SKU-77$59Yes1,880
SKU-08$15Yes9,540
FinanceWhich transactions are anomalous versus each account's own baseline?
TxnAvg 30dAmountAnomaly
TXN-41$30K$28K0.04
TXN-77$30K$1.5M0.96
TXN-90$49K$52K0.07
Contact

Build with us.

Bring us your data, or bring us your ambition. Run a proof of concept on your own tables, or join the team building the model — either way, we'd like to hear from you.

contact@nums.world
Proof of concept

Have a prediction problem sitting in a table? We'll run it through the model with you and show what a single inference pass can do.

Careers

We're looking for researchers and engineers who want to shape a foundation model for the world's tables. Tell us what you'd build.