Fundamental
Fundamental helps enterprises predict from tabular data instantly, no preprocessing required.
Fundamental builds Nexus, a Large Tabular Model (LTM) purpose-built for structured data analytics that delivers accurate predictions from raw tabular data with minimal configuration. Unlike traditional ML requiring months of preprocessing and data science work, Nexus learns complex feature interactions automatically, often through a single line of code. The company serves Fortune 100 enterprises using it for demand forecasting, price optimization, fraud detection, and customer churn prediction.
Problem solved
Traditional tabular ML methods like XGBoost require months of specialized data science work, extensive preprocessing, and break on missing values, while achieving inconsistent accuracy.
Target customer
Fortune 100 enterprises and mid-market companies with large structured datasets requiring predictive analytics, demand forecasting, and revenue optimization.
Founders
J
Jeremy Fraenkel
CEO & Co-Founder
Co-founder of Drift, previously at JP Morgan and Bridgewater, with two successful exits; Master's in Machine Learning from UC Berkeley.
M
Marta Garnelo
Chief Science Officer & Co-Founder
DeepMind alumna with PhD in Machine Learning and AI from Imperial College London; prior roles at Imperial College London and Instituto Superior Técnico.
G
Gabriel Suissa
Co-Founder
Israeli entrepreneur leading R&D operations in Israel with team members from AI21 Labs.
A
Annie Lamont
Co-Founder
Co-founder and Managing Partner at Oak HC/FT, lead investor in Fundamental's Series A.
Funding history
Seed
$30M
October 2024
Led by Unknown
· Unknown
Series A
$225M
February 5, 2026
Led by Oak HC/FT
· Valor Equity Partners, Battery Ventures, Salesforce Ventures, Hetz Ventures, with angels including Aravind Srinivas (Perplexity), Assaf Rappaport (Wiz), Henrique Dubugras (Brex), Olivier Pomel (Datadog)
Total raised:
$255M
Industries
Pricing
Enterprise contact-based pricing. Not publicly available. Available via AWS dashboard for AWS customers.
Notable customers
Multiple Fortune 100 companies with seven-figure contracts. Specific names not disclosed publicly.
Integrations
AWS (strategic partnership with dashboard integration), Cloudera (Enterprise AI Ecosystem partnership)
Tech stack
React (JavaScript frameworks)
HTTP/3
Linkedin Insight Tag (Analytics)
Google Analytics (Analytics)
HSTS (Security)
Apple iCloud Mail (Webmail)
Google Workspace (Email)
Cloudflare (CDN)
HubSpot (Marketing automation)
DoubleClick Floodlight
Google Tag Manager (Tag managers)
Framer Sites (Page builders)
Website
Competitors
XGBoost & Traditional ML
Requires extensive preprocessing, fails on missing data, demands months of specialized data science work per use case.
Microsoft Excel Copilot / Google Sheets AI
Wraps LLMs around spreadsheet interfaces; Fundamental's native tabular architecture provides superior reasoning without relying on transformer models.
Google BigQuery ML / AWS SageMaker AutoML
General-purpose AutoML platforms; Fundamental's architecture is specifically optimized for tabular reasoning and deterministic outputs.
Why this matters: Fundamental's $255M Series A (one of the largest for a stealth AI startup) and DeepMind-caliber founding team signal serious institutional confidence in a novel approach to enterprise analytics. In a crowded AI market, Fundamental's purpose-built tabular architecture and deterministic outputs address a real pain point—traditional ML requires enormous data science resources—while being differentiated from shallow LLM wrappers around spreadsheets.
Best for: Enterprise data teams and analytics leaders needing fast, accurate predictions from structured data without extensive data science resources or preprocessing overhead.
Use cases
Demand Forecasting
Enterprises connect historical sales and operational data to Nexus to forecast customer demand across regions and time periods. The model automatically identifies seasonal patterns, feature interactions, and edge cases without manual feature engineering, reducing forecast error and improving inventory planning.
Price Optimization
Retailers and SaaS companies use Nexus to predict optimal pricing by analyzing competitor pricing, customer segments, demand elasticity, and historical conversion data. The model captures non-linear interactions between pricing and demand that traditional regression misses.
Fraud Detection & Financial Risk
Financial institutions connect transaction, account, and behavioral data to detect fraudulent patterns and assess credit risk. Nexus learns complex feature dependencies across millions of transactions without retraining, improving detection accuracy while reducing false positives.
Customer Churn Prediction
Enterprise SaaS and subscription companies predict which customers are likely to churn by analyzing usage patterns, billing data, and engagement metrics. Nexus provides actionable predictions enabling proactive retention campaigns.
Hospital Readmissions Prediction
Healthcare systems use Nexus to identify high-risk patients likely to be readmitted using patient history, treatment data, and demographic information. This enables targeted interventions and improves population health outcomes.
Alternatives
H2O AutoML
General-purpose AutoML platform; Fundamental is purpose-built for tabular data with deterministic, non-transformer architecture.
Dataiku
Broader data science and ML governance platform; Fundamental is narrowly focused on tabular prediction with minimal configuration.
AWS SageMaker Autopilot
Cloud-native AutoML requiring more setup and infrastructure knowledge; Fundamental offers single-line-of-code simplicity.
FAQ
What does Fundamental do? +
Fundamental builds Nexus, a Large Tabular Model (LTM) specifically architected for structured data. It delivers accurate predictions from raw tabular data with minimal configuration—often just a single line of code—automatically learning complex feature interactions without requiring preprocessing, fine-tuning, or retraining.
How much does Fundamental cost? +
Pricing is not publicly available. Fundamental uses an enterprise sales model. Contact sales at fundamental.tech/contact for pricing. AWS customers can access Nexus through their AWS dashboard.
What are alternatives to Fundamental? +
XGBoost and traditional ML require extensive preprocessing; H2O AutoML and Dataiku are broader AutoML platforms; AWS SageMaker Autopilot is cloud-native but requires more setup. Fundamental differentiates through purpose-built tabular architecture and minimal configuration.
Who uses Fundamental? +
Multiple Fortune 100 companies are using Nexus for demand forecasting, price optimization, fraud detection, and customer churn analysis. Target customers are large enterprises with substantial structured data and analytics teams.
How does Fundamental compare to XGBoost? +
XGBoost requires months of data science work, extensive preprocessing, and custom feature engineering for each use case. Fundamental's Nexus delivers predictions automatically from raw data with a single line of code, without preprocessing or retraining, often achieving higher accuracy.
Is Fundamental's model deterministic? +
Yes. Unlike LLM-based approaches that produce variable outputs, Nexus is deterministic—it returns the same answer every time for the same input. It does not rely on transformer architecture, distinguishing it from contemporary AI approaches.
Tags
tabular data
foundation models
predictive analytics
enterprise AI
demand forecasting
price optimization
fraud detection
no-code ML