Berlin Tech Meetup: The Future of Relational Foundation Models, Systems, and Real-World Applications
For Research Scientists
The accuracy gains are real and reproducible. KumoRFM learns from relational structure that flat feature tables destroy. It's also pre-trained on thousands of proprietary and public datasets, providing additional pattern knowledge that delivers 10-50% accuracy gains over baselines on Stanford RelBench. The methodology is published at NeurIPS, ICML, and KDD. Every claim is verifiable.
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 Research Scientists choose Kumo
Evaluate Kumo the way you evaluate any research — with papers, benchmarks, and reproducible experiments.
40+
Peer-reviewed publications
NeurIPS, ICML, KDD. The methodology behind KumoRFM is published, cited, and advancing the field. No black-box vendor claims.
RelBench
Open benchmark
Evaluated on Stanford's RelBench — 11 databases, 30 tasks. Results are public and reproducible. Compare against any method.
Reproducible
Experiments on your data
Run KumoRFM on your own relational data and compare against your existing baselines. The evaluation framework is transparent.
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 — Most vendor AI is opaque: No published methodology, no public benchmarks, no way to verify claims independently. You are asked to trust marketing materials instead of peer-reviewed research. For a scientist, that is a non-starter.
Path 2 — LLMs for structured data: They tokenize relational data as text, destroying the structural information that encodes predictive signal. There is no theoretical grounding for why this should work on relational databases — and empirically, it does not.
Kumo is built on a transparent, benchmarked, and peer-reviewed foundation. Every method is published, every result is reproducible via RelBench, and the theoretical grounding is rooted in 40+ papers at the world's top ML venues.
40+
Peer-reviewed papers
Published methodology
NeurIPS/ICML/KDD
Top-tier venues
Rigorous peer review
Open
Benchmarks (RelBench)
Reproducible evaluation
100%
Reproducible methodology
Public and verifiable
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
Our methodology is fully open. Evaluate the science behind Kumo through 40+ peer-reviewed papers at NeurIPS, ICML, and KDD — all 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
