Simple, Transparent Pricing

Yes, you can upgrade or downgrade your plan at any time. When upgrading, the new features are available immediately. When downgrading, changes take effect at the end of your current billing cycle.

You'll receive a notification when you reach 80% of your limit. Once reached, workflows will pause until the next billing cycle or until you upgrade your plan.

Yes! NodeLoom can be self-hosted on your own infrastructure with a one-time license key. This gives you full control over your data and unlimited executions. Self-hosted deployment is available on the Enterprise plan.

NodeLoom supports OpenAI, Anthropic (Claude), Google Gemini, and Ollama for local models. You bring your own API keys, and we never access your AI credentials.

Both the Team and Business plans include a 14-day free trial so you can evaluate the platform with your team. No commitment required. For Enterprise, we offer personalized demos tailored to your use case.

Absolutely. All credentials are encrypted at rest and in transit using industry-standard encryption. Code execution runs in fully isolated sandboxes. We provide RBAC, comprehensive audit logging, and enterprise-grade security controls.

Yes. Our observability SDKs (Python, TypeScript, Java, Go) let you instrument any AI agent, whether built with LangChain, CrewAI, or a custom framework. SDK-instrumented agents get the same monitoring, drift detection, and compliance tracking as native workflows.

Agent Discovery automatically finds AI agents running across your infrastructure. It scans cloud providers (AWS, GCP, Azure), GitHub repositories, and MCP gateways to build a complete inventory of your AI agents, even ones your team didn't know existed. Available on Business and Enterprise plans.

Team plans include email support. Business plans get priority support with 8-hour response times. Enterprise plans include a dedicated account manager with a dedicated Slack channel.

Business plans include LLM-as-Judge evaluation for scoring agent outputs against custom criteria. Enterprise plans add red team adversarial testing, which runs automated attacks (prompt injection, jailbreak, data exfiltration) against your AI agents and generates vulnerability reports.

Business and Enterprise plans include automated incident response playbooks that trigger when drift thresholds or guardrail violations are detected. Playbooks can quarantine affected workflows, notify team members via email or Slack, escalate to on-call staff, and automatically rollback to the last known-good version.

AI agent governance is the practice of establishing controls, policies, and oversight mechanisms for AI agents operating in production environments. It encompasses monitoring agent behavior, enforcing guardrails on inputs and outputs, maintaining audit trails, ensuring compliance with regulations, and testing agents for security vulnerabilities. NodeLoom provides a comprehensive AI agent governance platform that covers discovery, monitoring, guardrails, compliance automation, and adversarial testing.

AI observability goes beyond traditional monitoring by providing deep visibility into how AI agents behave, reason, and make decisions. It includes tracking token usage, monitoring output quality through LLM-as-judge evaluation, detecting behavioral drift over time, analyzing sentiment in responses, and tracing the full execution path of agent workflows. NodeLoom provides AI observability through SDKs for Python, TypeScript, Java, and Go, plus zero-instrumentation eBPF probes for kernel-level monitoring.

Monitoring AI agents in production requires instrumenting agents with observability SDKs, establishing behavioral baselines, detecting anomalies and drift, tracking token usage and costs, and setting up alerting. NodeLoom automates this with lightweight SDKs that add less than 1ms overhead, automatic baseline learning, configurable anomaly and drift thresholds, and incident response playbooks that trigger automated actions when issues are detected.

AI guardrails are safety controls that validate AI agent inputs and outputs to prevent harmful, non-compliant, or unintended behavior. Types of guardrails include keyword and regex filtering, LLM-as-judge evaluation (where a separate AI scores outputs on dimensions like safety and accuracy), semantic similarity detection (comparing against known-bad patterns), PII redaction, prompt injection detection, and custom rules. NodeLoom supports all these guardrail types with configurable severity levels (WARNING, BLOCK, LOG).

Shadow AI refers to AI agents deployed by teams without centralized visibility or governance. NodeLoom Agent Discovery automatically finds these agents by scanning cloud providers (AWS, GCP, Azure), GitHub repositories, container orchestrators, and MCP gateways. For Linux hosts, an optional eBPF kernel-level agent intercepts LLM API calls at the TLS layer, detecting any process communicating with OpenAI, Anthropic, Google, or other LLM providers — without requiring code changes.

AI monitoring focuses on observability: tracking what AI agents are doing, detecting anomalies, and alerting on issues. AI governance is broader — it includes monitoring plus enforcement of policies, compliance automation, access controls, audit trails, adversarial testing, and incident response. NodeLoom is an AI governance platform that includes monitoring as a core capability alongside guardrails, compliance dashboards, red team testing, RBAC, and automated playbooks.

NodeLoom offers full self-hosted deployment on your own infrastructure using Docker Compose, Kubernetes, or Helm charts. Self-hosted deployments include all platform features (monitoring, guardrails, compliance, red team testing) and support air-gapped environments with no external data transmission. This is critical for organizations in regulated industries (healthcare, financial services, defense) that require data sovereignty.

NodeLoom provides automated red team adversarial testing. An attacker LLM generates targeted attacks across 6 categories — prompt injection, jailbreak, data exfiltration, tool abuse, PII leakage, and harmful output. Each attack is sent to the target AI agent, and a judge LLM evaluates whether the attack succeeded. Results include a resilience score (1.00-5.00), per-finding severity ratings, and remediation recommendations. Critical findings automatically trigger incident response playbooks.

NodeLoom supports compliance reporting for SOC 2, HIPAA, GDPR, ISO 42001, NIST AI RMF, and PCI-DSS. The platform provides a compliance dashboard showing guardrail coverage, audit trail integrity, credential health, and user access patterns. One-click report generation creates framework-specific compliance reports. The cryptographic audit trail uses SHA-256 hash chaining for tamper-proof event logs.

Yes. NodeLoom SDKs include built-in integrations for LangChain and CrewAI. The Python SDK provides a LangChain callback handler and CrewAI decorator that automatically capture traces, spans, token usage, and tool calls. The TypeScript SDK includes a LangChain callback handler. Any framework can be instrumented manually using the trace/span API. SDK-instrumented agents get full monitoring, drift detection, evaluation, and compliance tracking.