Fundamental’s Large Tabular Model NEXUS is now available on Amazon SageMaker JumpStart

2026-06-03 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Data Science & Analytics · Depth: Intermediate, medium

Summary

Fundamental's NEXUS, a large tabular foundation model, is now available on Amazon SageMaker JumpStart, enabling enterprises to generate accurate, deterministic predictions from structured data rapidly. Pre-trained on billions of real-world prediction tasks, NEXUS offers a deterministic architecture for consistent results, native tabular understanding without manual feature engineering, and non-sequential reasoning for multi-dimensional relationships in tables. Unlike traditional ML, which can take 3-6 months, or non-deterministic LLMs, NEXUS processes massive datasets up to billion-row capability, handles cross-schema reasoning, and performs autonomous data cleaning. It deploys on a dedicated "ml.p5en.48xlarge" instance with 8x NVIDIA H200 GPUs via AWS Marketplace, integrating with the Fundamental Python SDK for training and prediction, with data remaining secure within the user's AWS environment. This solution supports diverse use cases like fraud detection, patient risk stratification, and demand forecasting.

Key takeaway

For Data Scientists or ML Engineers building predictive models on structured enterprise data, you should evaluate Fundamental's NEXUS on Amazon SageMaker AI. This solution accelerates time-to-value by providing a pre-trained, deterministic tabular foundation model that eliminates extensive feature engineering and model training, which typically takes months. Consider deploying NEXUS for critical business decisions requiring scalable, compliant, and consistent predictions from your tabular datasets.

Key insights

NEXUS is a foundation model purpose-built for tabular data, delivering deterministic, scalable predictions without extensive feature engineering.

Principles

• Tabular data benefits from specialized foundation models.
• Deterministic outputs are critical for enterprise prediction tasks.
• Pre-training on diverse tasks reduces manual feature engineering.

Method

Deploy NEXUS from AWS Marketplace to a SageMaker AI endpoint, install the Fundamental Python SDK, upload data to Amazon S3, then use "clf.fit()" and "clf.predict()" for training and inference.

In practice

• Access NEXUS base or industry variants on SageMaker JumpStart.
• Utilize the Fundamental Python SDK for familiar ML API.
• Deploy on SageMaker AI for managed, compliant infrastructure.

Topics

• Large Tabular Models
• Amazon SageMaker
• Foundation Models
• Tabular Data Prediction
• Enterprise AI
• Deterministic AI

Code references

• Fundamental-Technologies/fundamental-cookbook

Best for: Machine Learning Engineer, Data Scientist, Director of AI/ML

Related on AIssential

• Beyond the lakehouse: Fundamental's NEXUS bypasses manual ETL with a native foundation model for tabular data — NEXUS is a Large Tabular Model designed to natively understand and predict from complex, non-sequential business data.
• Agent-guided workflows to accelerate model customization in Amazon SageMaker AI — Agent-guided workflows in SageMaker AI simplify foundation model customization from prompt to deployment.
• Is XGBoost gone: How Relational Foundation Models Conquered 500 Billion Row Enterprise Data — The dominance of gradient boosted trees on tabular data was an illusion created by pre-processed Kaggle datasets.
• TabH2O: A Unified Foundation Model for Tabular Prediction — TabH2O unifies tabular classification and regression via in-context learning, improving training stability and noise robustness.
• When Tabular Foundation Models Meet Strategic Tabular Data: A Prior Alignment Approach — Strategic manipulation creates a prior mismatch for tabular foundation models, which SPN corrects at inference.
• TabPFN AI Accelerates Business Transformation on Databricks — TabPFN brings LLM-like pre-trained efficiency to tabular data, dramatically simplifying traditional ML workflows.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.