Palma.ai
Palma.ai provides an enterprise governance layer for Model Context Protocol (MCP) and AI agent infrastructure, enabling organizations to control which tools and Skills agents can access. The platform enforces identity-based entitlements, evaluates policies on actual call arguments, brokers credentials per call, and maintains a tamper-evident audit trail—all through a single governed connector compatible with major AI assistants and deployment options including VPC or air-gapped environments.
- Artificial Intelligence
- Cybersecurity
- Enterprise Software
- Developer Tools
- Software Only
- AI Agents
Funding
Raised to date
$1.8MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 1 round
AIDDSN$1.8M · Pre-Seed
Announced Sep 2026 · from article date
Disclosed cumulative: $1.8M
Founders
Product
Problem
Enterprises deploying AI agents across multiple developer tools face challenges in routing calls, enforcing access policies, and monitoring performance, cost, and security. Without a centralized control layer, MCP (Model Context Protocol) teams must manage server reliability manually while agent teams struggle to ensure consistent behavior across heterogeneous client environments.
Solution
Palma.ai provides a unified platform that sits between MCP servers and client tools such as VS Code, Cursor, and Copilot. It handles request routing, applies configurable policies, and offers real-time observability for each tool call. The platform resolves every call to a specific person via identity provider groups, evaluates policies on the actual arguments of each call, brokers credentials from the enterprise vault on a per-call basis, and writes a tamper-evident audit record for every decision. By grouping tools by capability and tracking metrics for cost, security, and performance, the platform lets MCP teams concentrate on server design while agent teams focus on building robust AI agents.
Target Audience
Primary customers are enterprise security, platform, and MCP infrastructure teams responsible for tool governance, as well as AI agent development teams that need a controlled, observable environment for deploying agents across developer tools.
Features
- Central routing layer that directs AI agent calls from any MCP-enabled client to the appropriate server through a single governed connector
- Identity-based entitlement: access follows IdP groups (Entra, Okta, or any OIDC provider), so users get the same tools and Skills in any assistant they open
- Policy engine that evaluates rules on the actual arguments of each call—allow, hold for a named approver, or block—including support for cost limits and thresholds
- Pre-policy argument scanning for secrets, personal data, and injected instructions
- Per-call credential brokering from the enterprise vault; end users never hold credentials directly
- Tamper-evident, offline-verifiable audit trail with cost attribution per team, streaming to existing SIEM systems
- Integration with all major MCP-enabled clients including Claude, VS Code, ChatGPT, JetBrains, Gemini, Copilot Studio, and Cursor
- Deployment options include VPC or fully air-gapped on-premises environments