About EntityFold

EntityFold is exploring the infrastructure needed to build intelligent systems around entities, relationships, and evolving data.

Modern applications increasingly need to make real-time decisions about entities such as users, organizations, transactions, and devices. But the data and infrastructure behind these decisions is often scattered across databases, caches, streaming systems, and batch pipelines.

EntityFold explores how these pieces can come together into a more unified foundation for building real-time, entity-aware intelligent systems.

Areas of Focus

Our ongoing engineering efforts center on three core problems:

• Entity-Centric Modeling: Treating entities and their relations as primary schema primitives rather than isolated rows.
• Point-in-Time Correctness: Guaranteeing deterministic, mathematically sound historical state reconstruction without data leakage.
• Unified Execution Runtimes: Eliminating training-serving skew by compiling single declarative schemas into both streaming and analytical engines.

Writing & Progress

We are writing about the ideas, systems, and engineering challenges involved in building this kind of infrastructure.

To follow along, read our latest write-ups on the Notes page.