Berlin Tech Meetup: The Future of Relational Foundation Models, Systems, and Real-World Applications
For Data Scientists
Feature engineering doesn't just waste your time. It limits accuracy by flattening relational data into feature tables that lose multi-hop signals. KumoRFM learns directly from your relational schema. It's also pre-trained on thousands of datasets, discovering patterns you'd never think to engineer, delivering an additional accuracy boost of 10-50% on RelBench. And it produces production models in hours, not months.
Book a demo and get a free trial of the full platform: research agent, fine-tune capabilities, and forward-deployed engineer support.
By submitting, you accept the Terms and Privacy Policy.
Why Data Scientists choose Kumo
You became a data scientist to solve problems, not wrangle data. Kumo gives you that back.
Minutes
To your first prediction
Write a prediction query in Kumo's PQL. Get a production-ready model with evaluation metrics, explainability, and continuous retraining — in minutes, not months.
10–50%
Over your current models
KumoRFM learns from relational structure that flat feature tables miss. Beat your best XGBoost model without a single line of feature engineering.
Built-in
Explainability & evaluation
Every prediction comes with feature importance scores and evaluation metrics. Understand what drives the model — no separate explainability tooling needed.
Loved by data scientists, ML engineers & CXOs at
In production today
Enterprise Customer
Went from 3 models in production to over 50 in a single quarter with the same team. Feature engineering eliminated, pipeline complexity reduced by 95%.
Databricks
Lead-scoring models delivering dramatic improvement in conversion rates. Deployed in days instead of months, with zero feature engineering.
DoorDash
Restaurant recommendations driving hundreds of millions in GMV. Expanded to notification reranking and send-time optimization using the same foundation.
The challenge you know too well
Path 1 — Feature engineering by hand: You spend 80% of your time writing SQL joins, temporal aggregations, and window functions — not building models. Every new prediction task means weeks of pipeline work before you can even train a baseline.
Path 2 — Try LLMs on your data: Large language models flatten your relational data into text. They have no understanding of primary keys, foreign keys, or the multi-hop relationships between tables — exactly the structure that makes your predictions accurate.
Kumo's Predictive Query Language lets you express what you want to predict in a single query. The foundation model learns directly from your relational schema — no feature engineering, no flattening, no pipeline code. You get production-ready models with built-in explainability and evaluation metrics.
95%
Less feature engineering
Go from query to model directly
10–50%
Accuracy boost
Over traditional ML (RelBench)
Minutes
To first prediction
Not weeks of pipeline work
Built-in
Explainability
Feature importance out of the box
KumoRFM isn't limited to your data alone. Pre-trained on billions of relational patterns across diverse datasets and fine-tuned to your schema, it sees what no in-house model can. As per the SAP SALT benchmark.
LLM
GPT4 + AutoML
PhD Data Scientist
Feature eng. + XGBoost
KumoRFM
Relational Foundation Model
17x
increase in models shipped per quarter
“Beating internal XGBoost model on key metrics with far less data and features. We went from three models in production to over fifty in a single quarter, with the same team.”
Matt Loskamp
GTM Data Science Leader, Enterprise Customer
From startups to enterprises, leading organizations rely on Kumo to deliver predictive insights at scale.
Peer-reviewed
Kumo is built on 40+ peer-reviewed papers at NeurIPS, ICML, and KDD. The methodology is public and reproducible.
K. Huang, M. Fey, J. Leskovec et al.
A foundation model for relational data - pre-trained across schemas, it delivers accurate predictions out of the box and improves with fine-tuning on your specific data.
Read paperM. Fey, W. Hu, K. Huang, J. Leskovec et al.
Introduces learning predictive models directly on relational databases, eliminating the feature engineering pipeline that has historically bottlenecked enterprise ML.
Read paperJ. Robinson, R. Miao, K. Huang et al.
An open benchmark for evaluating relational prediction methods across 11 databases and 30 tasks. Kumo consistently outperforms traditional ML baselines.
Read paper
