RESEARCH

Research-first approach to data infrastructure

Velum is built by researchers applying academic rigor to enterprise data problems. We believe the gap between raw data and useful intelligence requires new foundations—not just better tooling.

Backed by Y Combinator (W26).

KEY CONCEPTS

The ideas we build on

Ontology

A formal specification of a domain—the entities, relationships, and rules that define what things mean. Think of it as a schema with semantics.

Data Mesh

A decentralized approach to data architecture where domain teams own their data as products, with contracts ensuring interoperability across the organization.

Knowledge Hypergraph

A data structure where relationships can connect any number of entities, not just pairs. Better suited for modeling complex operational processes.

TEAM

Researchers building for enterprise

We're researchers with backgrounds across knowledge representation, distributed systems, and machine learning—now building the semantic infrastructure that modern data platforms need. Backed by Y Combinator (W26).

FOUNDERS

Alen Rubilar-Muñoz

Co-founder

Benjamin Muñoz-Cerro

Co-founder

RESEARCH AFFILIATIONS

Harvard University Stanford University Vanderbilt University Leipzig University Minerva University Max Planck Institute for Intelligent Systems Max Planck Institute for Software Systems Meiler Lab

RESEARCH DIRECTIONS

What we're exploring

We're working at the intersection of knowledge representation, NLP, and data engineering.

01Knowledge representation for enterprise data

02Semantic extraction from unstructured sources

03Ontology-driven data contracts and governance

04Scalable knowledge graph construction

We publish what we learn. Papers coming soon.

Work with us

We're looking for design partners with interesting data problems and researchers interested in knowledge representation, NLP, and distributed systems.