Baffle

Baffle is a Santa Clara-based data-security company built around a simple thesis: if data is encrypted, tokenized, or masked at the field level before it ever reaches a database, warehouse, or AI pipeline, a breach of that downstream system doesn’t expose usable data. Founded in 2015 by Ameesh Divatia and Priyadarshan Kolte, the company markets this as Encryption-as-a-Service — a proxy-based approach that adds protection without requiring applications to be rewritten to handle cryptographic operations directly.

The company has raised $36.5M across multiple rounds, including a Series B led by Celesta Capital with participation from National Grid Partners, Lytical Ventures, and Nepenthe Capital — a real institutional round, though modest compared to the largest data-security platform raises. Baffle has increasingly positioned its tokenization and encryption approach for AI and data-pipeline use cases, where sensitive data moving into training sets or analytics platforms creates similar exposure to what databases faced a decade ago.

For a CISO, the pitch is straightforward: field-level encryption reduces breach blast radius and can simplify compliance scope for PCI, HIPAA, and similar regimes by keeping sensitive fields protected end-to-end. The tradeoff to evaluate is operational: proxy-based encryption architectures add a component in the data path, and prospective customers should look for named production deployments and performance benchmarks rather than relying on the "breaches irrelevant" framing in isolation.