Velum Labs

Velum is an open-source firewall for content-level access control across documents, databases, and applications. Admins define policies, and we apply them in real time across your entire stack to control what each recipient sees.

Our firewall sits between your data sources and consumers (LLMs, users, vendors), enforcing policies on every request. We understand meaning in text, documents, and databases to detect sensitive information based on your policies, then apply recipient-specific controls without blocking workflows.

Create policies in natural language. Teams write rules that define what's sensitive and who should access it, and Velum enforces them consistently across all systems. When authorization changes, we can reveal tokenized data just-in-time for approved recipients, maintaining both security and utility.

We integrate directly into your existing infrastructure. For AI workflows, we intercept prompts, responses, and retrievals to prevent sensitive information from leaking into or out of models. For enterprise systems like Salesforce, SAP, or Workday, we extend policy enforcement to exports, reports, and syncs.

Velum is designed for scale and compliance. We provide encryption and structure-preserving tokenization for high-throughput pipelines, with multi-cloud deployment options and configurable data residency per tenant to meet your compliance requirements.

High-Level Overview

Velum Labs is a San Francisco-based startup founded in 2025 that builds an open-source dynamic data firewall designed to control access to sensitive information across both AI systems and human users. Their product operates at the content level, applying real-time policy enforcement across documents, databases, and applications to prevent unauthorized data exposure. Velum’s firewall integrates seamlessly into existing enterprise infrastructure, intercepting data flows to large language models (LLMs), APIs, and operational systems like Salesforce or SAP, ensuring sensitive information is redacted or tokenized according to customizable, natural language policies. This enables organizations to maintain compliance and security while preserving data utility and workflow efficiency.

Velum primarily serves enterprises that require stringent data governance in AI and human workflows, addressing critical challenges around data privacy, regulatory compliance, and secure AI adoption. The company’s solution is particularly relevant as organizations increasingly deploy AI tools that interact with sensitive unstructured data. Velum’s growth momentum is underscored by its acceptance into the Y Combinator Fall 2025 batch and its positioning as a foundational security layer for AI-driven enterprises[1][2][3].

Origin Story

Velum Labs was founded in 2025 by Benjamin Muñoz-Cerro (CEO) and Alen Rubilar-Muñoz (CTO), both with strong academic backgrounds—Benjamin in quantum computing at Stanford and physics at Harvard, and Alen as a machine learning software engineer and mathematician. The idea emerged from the need to secure sensitive information in the rapidly evolving AI landscape, where traditional perimeter security fails to address data exposure risks within AI workflows and human interactions. Early traction came through Y Combinator’s Fall 2025 batch, providing validation and initial support for their open-source firewall approach[1][3].

Core Differentiators

• Content-Level Access Control: Velum enforces policies at the granularity of sentences, tables, and even figure regions, allowing precise redaction or tokenization rather than coarse blocking.
• Natural Language Policy Engine: Teams write access rules in natural language, simplifying policy creation and management.
• Real-Time Enforcement: The firewall operates in-path, intercepting every data request to apply policies dynamically across all data consumers, including AI models and human users.
• Integration with AI and Enterprise Systems: Velum supports AI workflows by intercepting prompts and responses to prevent data leakage and extends policy enforcement to enterprise platforms like Salesforce, SAP, and Workday.
• Compliance-Ready Architecture: Features include encryption, structure-preserving tokenization, multi-cloud deployment, configurable data residency, and audit-ready logs.
• Open-Source Model: Encourages transparency, extensibility, and community collaboration[1][2].

Role in the Broader Tech Landscape

Velum Labs is riding the critical trend of data privacy and security in AI adoption, addressing a growing market need as enterprises integrate AI tools that handle sensitive unstructured data. The timing is crucial because traditional security models do not adequately protect data flowing into and out of AI systems, creating compliance and leakage risks. Market forces such as increasing regulatory scrutiny (e.g., GDPR, HIPAA), rising AI adoption, and demand for explainable, secure AI workflows favor Velum’s approach. By providing a dynamic firewall that enforces least-privilege access in real time, Velum influences the broader ecosystem by enabling safer AI deployment and fostering trust in AI-human data interactions[1][2].

Quick Take & Future Outlook

Velum Labs is well-positioned to become a foundational security layer for enterprises deploying AI at scale, especially in regulated industries. Going forward, the company is likely to expand its integrations, enhance policy automation capabilities, and scale its platform to support more complex AI workflows and data types. Trends such as increasing AI regulation, demand for privacy-preserving AI, and the rise of multi-cloud environments will shape Velum’s journey. Its open-source model may also foster a growing community that drives innovation and adoption. Velum’s influence will likely grow as organizations seek to balance AI utility with stringent data governance, making it a key player in the evolving AI security landscape[1][2][3].