Architectural Overview

Fount's Architecture

Fount is built as a unified causal model designed to learn how different parts of a business system interact over time.

It is powered by four core technologies working together:

• Causal AI: Identifies true cause-and-effect relationships across your data, moving beyond correlation to understand which factors actually drive outcomes. Built on Judea Pearl's causal inference framework, it models interventions and answers "what if" questions about scenarios that haven't happened yet.
• Deep Learning: Captures complex, non-linear relationships and interconnected impacts across large datasets, handling multiple KPIs and intricate variable dependencies simultaneously.
• Multivariate Forecast: Simultaneously forecast multiple interconnected variables to identify the best possible outcomes across the entire business system, not just individual parts.

Rather than modeling each variable in isolation, Fount learns relationships across all inputs and outputs simultaneously. These include factors such as pricing, promotions, media activity, and external conditions, along with outcomes like demand, revenue, and market share.

By modeling everything within a single system, Fount captures how changes in one variable influence others. For example, a change in pricing may impact demand, which in turn affects revenue and inventory. These relationships are learned directly, not approximated through separate models.

This approach enables Fount to move beyond isolated predictions and provide a more complete understanding of the underlying system.

Core Capabilities