Natoma

Overview

Natoma focuses on non-human identity governance, including identities for AI agents, helping teams discover, manage, and secure machine and agent access. Founded in 2023 in San Francisco, it is aimed squarely at the emerging agentic wave, where autonomous agents need identities, scopes, and guardrails. It is an early-stage vendor, so treat references and scale claims cautiously.

Snowflake signed a definitive agreement to acquire Natoma on 27 May 2026 for roughly 110 million dollars, mostly in stock. Natoma's governed Model Context Protocol server catalogue is destined for Cortex Agents and Snowflake Intelligence, so it is unlikely to remain a neutral cross-platform product. If you are evaluating outside the Snowflake estate, weigh the alternatives in our AI agent identity ranking and read the news item.

What it is good at

Its focus on agent and machine identity governance is timely: it provides discovery of NHIs and AI agents plus policy controls to bound what they can access. Governance, lifecycle, and authorization are the strongest dimensions. For teams piloting AI agents that call internal tools and APIs, having guardrails and an inventory designed for that pattern is a real advantage over retrofitting human-centric IGA.

Where it falls short

As a 2023-founded startup, scale, references, and proven enterprise hardening are limited, so pilot before committing. It is not a human IGA tool, so workforce certification and access reviews for employees sit outside its core. It is also not an identity provider, so authentication, SSO, and MFA are out of scope. Agent identity standards are still forming, which means the roadmap will shift.

Pricing

No public pricing. Quote-based subscription through a sales-led process. Model NHI and agent counts with the TCO calculator before evaluating.

Best for, and who should look elsewhere

A fit for teams governing non-human identities and AI agents that want purpose-built guardrails early. Look elsewhere if you need a proven large-scale platform or only human IGA. Compare with Astrix Security, Token Security, and the broader AI identity category.

Bottom line

An emerging, agent-focused governance option for non-human and AI-agent identities. Well-timed premise; validate maturity with a pilot.