What Fantix does

Fantix is a US-based, privately-held AI platform founded in 2022 (incorporated in New York) that builds privacy-preserving machine learning infrastructure for business and marketing data science. Its core technology is a federated learning stack that trains models across a network of large data owners without pooling user-level data — only hashed emails are transmitted and ML runs on partner premises — enabling collaborative AI development that satisfies GDPR/CCPA constraints and the post-cookie advertising environment. On top of this infrastructure, Fantix offers a portfolio of proprietary, trademarked products: Yellowcake (consumer attribute prediction with claimed 90%+ accuracy), Supernova (synthetic US consumer panels trained on 300M consumers and $3.5T in transactions from 2021 to present), LightID (cookieless probabilistic ID resolution identifying up to 90% of anonymous visitors), Fusion (federated lookalike audience builder pushed to programmatic marketplaces), Twinlytics (a natural-language AI agent for market scenario exploration), plus customization offerings (Fine-tune and Models Training) and the AI Accelerator for agentic AI startups.

The company serves four primary customer segments: financial institutions and product/research teams using Supernova for synthetic market research; brands, e-commerce companies, and publishers using Yellowcake for CRM enrichment and Twinlytics for market exploration; advertisers and ad-tech partners using LightID and Fusion for cookieless targeting; and AI startups via the AI Accelerator. Revenue is generated through quote-based enterprise subscriptions (annual Order Forms) supplemented by usage-based API access, federated model-training services, and licensing/partnership arrangements with the AI Accelerator. Distribution combines direct enterprise sales (with a free 15-minute demo) and API-led self-serve access, supported by partnerships with LiveRamp, Magnite, OpenX, Meta, and Google for downstream activation, and data-contribution partnerships with organizations such as Sports Innovation Lab.

In February 2026, Life360 acquired Fantix's advertising unit (LightID and Fusion plus related integrations) as part of its ad-tech platform buildout, expanding its reach to over 95% of US ad-eligible adults, while Fantix continued operating its core AI platform independently. The company has raised $1.6M in a single October 2023 seed round from B Heroes, Connect Ventures, Fastweb, Founders Factory, Gaingels, IAG Capital Partners, and Notion Capital, and operates with 11–50 employees.

Name

Fantix

Legal name

Fantix, Inc.

Website

https://fantix.ai

Company type

Private

Founded year

2022

Operating status

Operating

Headcount range

11–50 employees

Short description

Fantix is a US-based AI platform that uses federated learning to deliver privacy-first synthetic data, consumer attribute prediction, and cookieless ID resolution for advertisers, brands, financial institutions, and AI startups without exposing underlying user data.

Ownership category

akta.pro rank

Product category

Privacy-Preserving AI Data Platform

NAICS

Software Publishers (5132)

SIC

Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

Confidential AI & Privacy-Preserving ML (federated learning, MPC, HE, TEEs) (HDAAAKAI)

akta.pro secondary industry

Federated Learning & On-Device Personalization (HDAAAJAG)

Keywords

• Federated machine learning
• Synthetic data generation
• Cookieless identity resolution
• Predictive consumer analytics
• Privacy-preserving AI platform

Headquarters

HQ city

Nyack

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Data and AI Services (SaaS Platform): Fantix licenses its proprietary AI platform, models, and synthetic data to enterprise clients. Services are provided via subscription and usage-based models, with clients paying for access to the Platform and Data as described in Order Forms. The platform includes API access to models like Yellowcake, synthetic user panels from Supernova, and federated audience building capabilities.
2. API Access and Usage: Clients query models via API and receive responses in real time. The usage-based component comes from API call volumes and data query requests, particularly for consumer enrichment services like Yellowcake.
3. Federated Learning Services: Training of custom models using Fantix's federated infrastructure. Clients can develop proprietary models while data remains secure on their premises. Revenue derived from infrastructure access, expertise support, and model training services.
4. AI Accelerator Program: Pre-seed capital and platform access provided to qualifying AI startups in exchange for partnership and potential long-term commercial relationships. Supports development of first MVP in exchange for access to data federation and AI capabilities.

Pricing tiers

Go-to-market motion1 record

Distribution channels3 records

Marketing channels6 records

Core offering

Fantix operates a federated machine learning platform that trains AI models across a federation of large data owners without pooling user-level data, enabling privacy-first business data science and predictive analytics. Its products include pre-trained generative AI models for consumer attribute prediction (Yellowcake), synthetic user panel generation (Supernova), cookieless ID resolution (LightID), federated audience building (Fusion), and a natural-language AI assistant for market insights (Twinlytics), licensed to enterprises via subscription and API access.

Product overview

Fantix is an AI platform for business data science organized as a platform-plus-modules architecture. The core platform consists of federated machine learning infrastructure enabling collaborative AI training without pooling user-level data. Key products include: Yellowcake™ for consumer data enrichment via a universal generative model predicting hundreds of attributes; Supernova™ for creating synthetic user panels for market research; LightID™ for cookieless ID resolution maintaining visitor addressability; Fusion™ for federated audience building and lookalike targeting; and Twinlytics™ (AI assistant) for natural language exploration of market insights. Customization is offered through Fine-tune (enhanced model precision) and Models Training (training from scratch). The Data Federation platform provides the underlying infrastructure spanning 300 million US consumers and $3.5 trillion in transaction data. The AI Accelerator program supports Agentic AI startups.

Differentiator

Problem solved

Functional benefit

Brands

• Yellowcake: AI model for synthetic data generation and consumer data enrichment, capable of predicting hundreds of consumer attributes without knowing the consumer's identity.
• Supernova
• LightID
• Fusion
• Twinlytics

Products and services

• Yellowcake™ A universal generative AI model that predicts hundreds of consumer attributes (brand affinity, purchase behavior, demographics) without knowing the consumer's identity. Trained on payment transactions, mobility data, and demographics with over 90% accuracy, accessible via API using anonymous RFM (recency, frequency, monetary) data for CRM enrichment and consumer data enhancement.
• Supernova™ Generates comprehensive panels of synthetic US consumers mimicking real purchasing behavior at SKU and merchant levels. Trained on 300 million US consumers and $3.5 trillion in consumer spending from 2021 to present, with weekly (soon daily) panel regeneration for market research, investment research, and AI model training.
• LightID™ A probabilistic anonymous identifier replacing cookie-based identifiers for cross-device and cross-platform recognition of anonymous visitors, maintaining addressability of up to 90% of visitors (including unlogged traffic) for personalized marketing, campaign measurement, and programmatic monetization.
• Fusion™ A federated audience builder that creates lookalike and advanced targeting audiences for advertisers. Processes hashed emails securely on partner premises using deep learning algorithms to identify similar users, then maps results to alternative IDs and pushes optimized audiences to programmatic marketplaces (LiveRamp, Magnite, OpenX, Meta, Google).
• Twinlytics™ An AI-native conversational agent that helps strategy, product, and investment teams explore market 'what ifs' using natural language. Powered by high-fidelity transaction and mobility data representing over 20% of the US population, surfacing insights on spending patterns, foot traffic, and regional shifts.
• Fine-tune A customization service that fine-tunes pre-trained Fantix models with a client's own data to enhance accuracy, flexibility, and cost efficiency. Once fine-tuned, the model becomes proprietary to the client, providing customized solutions and competitive advantage.
• Train a Model A federated machine learning service enabling businesses to train custom AI models from scratch using Fantix's federated infrastructure. Data never leaves the owner's premises; only model updates are shared with the central server. The resulting model and IP are owned by the client.
• Data Federation Federated machine learning infrastructure tapping into extensive US consumer data sources including demographics, mobility patterns, consumer spending transactions (SKU and merchant level, $3.5 trillion total), and sports/entertainment preferences across 300 million US consumers via data partners such as Sports Innovation Lab.
• Synthetic Data GenAI for tabular data platform that creates artificially generated datasets mimicking real-world data characteristics. Privacy-safe (100% no PII), accurate (debiased through multi-dataset training), frequently updated, and scalable without being bound by real-world data set size, for use in UX personalization, targeting, and product development.
• AI Accelerator A program supporting founders of Agentic AI startups by providing access to Fantix's data federation, LLM capabilities, synthetic data generation, and pre-seed capital in exchange for partnership and potential long-term commercial relationships, supporting development of first MVPs.

Quantifiable outcome

• Over 90% accuracy for Yellowcake consumer attribute prediction
• +3 more outcomes

Segments5 records

Ideal customer profiles5 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration5 records

AI capability5 records

Feature9 records

Partnerships

Five partnerships are on record, tiered core.

• Sports Innovation LabcoreStrategic or Co-development PartnerSports Innovation Lab contributes its fan data to the Fantix federation. The loyalty of sports fans often extends beyond sports, influencing preferences in entertainment, fashion, and lifestyle products. Integrating sports fan data allows models to uncover cross-sector trends and correlations.
• The NewcocoreStrategic or Co-development PartnerThe Newco trained LightID using Fantix's federated machine learning infrastructure and data partners. LightID is a probabilistic model capable of identifying anonymous visitors, boosting addressability to above 90%.
• LiveRampcoreTechnology or IntegrationIntegration with LiveRamp for pushing optimized audiences to programmatic marketplaces. Fusion generates alternative IDs that are sent to LiveRamp and ad exchanges for targeting.
• Meta (Facebook)coreTechnology or IntegrationFusion pushes optimized audiences to Meta and Google for targeting. Meta IDs generated through the platform enable lookalike audience campaigns on Facebook.
• GooglecoreTechnology or IntegrationAudiences pushed to Google for targeting campaigns via programmatic marketplaces and Google Ads integration.

Scale indicators4 records

Recent moves6 records

Expansion highlights5 records

Emerging players

• Duality Technologies: Duality Technologies provides privacy-enhanced data collaboration using homomorphic encryption and secure multiparty computation—technically adjacent to Fantix's federated learning approach. Both target regulated industries needing collaborative AI/analytics without exposing underlying sensitive data, though Duality focuses more on cryptographic techniques.
• Cape Privacy: Cape Privacy (now Cape) is an emerging platform offering encrypted AI and federated learning capabilities for collaborative model training on encrypted data—an emerging player overlapping with Fantix's federated ML infrastructure. Both compete for regulated enterprise and financial services buyers needing privacy-preserving model training workflows.
• Hightouch: Hightouch is a reverse-ETL and data activation platform pushing warehouse customer data to ad platforms and marketing tools—directly analogous to Fantix's Fusion audience-building capability. Hightouch is at a more mature stage with broader activation destinations, representing both a partner-class peer and an emerging alternative for the marketing use case.

Broad incumbents

• Databricks: Databricks is a broad incumbent data + AI platform with native federated learning capabilities and a synthetic data offering via its Mosaic AI suite. As an established leader with massive R&D resources, it represents the dominant platform risk to niche players like Fantix across data science, model training, and collaborative ML workflows.
• Snowflake: Snowflake is a broad incumbent data cloud platform increasingly adding AI/ML, synthetic data, and data clean room capabilities that overlap with Fantix's federated data federation. Its massive enterprise footprint makes it the principal incumbent against which Fantix must differentiate on privacy-first architecture and pre-trained consumer models.

Direct peers

• Gretel: Gretel provides a developer-focused synthetic data and privacy-enhancing technology platform with APIs for generating synthetic datasets—an emerging competitor to Fantix's Yellowcake and Supernova. Both target data scientists building AI/ML pipelines with privacy-safe training data, though Gretel emphasizes developer ergonomics while Fantix emphasizes pre-trained consumer models.
• LiveRamp: LiveRamp is the leading data collaboration and identity resolution platform, providing the same cookieless identity, audience activation, and data connectivity use cases that Fantix's LightID and Fusion address. Fantix pushes audiences directly to LiveRamp, making LiveRamp both a partner and the dominant incumbent Fantix competes against for identity-led marketing data science.
• TripleBlind: TripleBlind provides a privacy-preserving ML platform enabling federated learning and confidential computing across data silos—directly analogous to Fantix's federated ML infrastructure. Both compete for enterprise buyers needing to train models without moving raw data, with TripleBlind operating at more enterprise scale and broader use cases.
• Mostly AI: Mostly AI is a synthetic data platform generating privacy-safe tabular and structured datasets for analytics and AI training—directly comparable to Fantix's Supernova synthetic consumer panels. Both target financial services, insurance, and consumer enterprises needing synthetic data for research and model training without PII exposure.
• Decentriq: Decentriq operates a confidential computing platform using secure enclaves for clean rooms and privacy-preserving data collaboration—directly overlapping with Fantix's federated learning approach to data collaboration without pooling. Both serve enterprise use cases requiring multi-party data analysis without exposing raw inputs.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/fantix

Revenue estimate

Valuation estimate

Number of profiles

Funding overview

Funding rounds1 record

Investors7 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Fantix