Nimble's data boundaries

Nimble provides web search and retrieval for AI agents, along with structured data from the public web for analytics teams. Nimble's platform is built on strict boundaries around what data is collected and how. This page explains those boundaries so customers can confirm Nimble is the right choice for their workload.

Public web data only

Nimble is engineered to collect information that is publicly accessible without login. That scope is intentional and part of the platform's compliance-by-design posture (SOC 2, GDPR, CCPA alignment, ethical IP sourcing). See the Trust Center for governance commitments.

Not in scope by design:

• Login-protected pages, private databases, and content requiring credentials, subscriptions, paywalls, or MFA

• Direct collection of sensitive personal information that would conflict with privacy regulations

• Activities that undermine site integrity (abusive request patterns, denial-of-service behavior)

• Uses intended to bypass explicit technical restrictions in ways that violate site terms or applicable law

Nimble operates primarily as a data processor on behalf of customers; the platform does not intentionally process personal data (Privacy Policy).

When Nimble is the right choice

Nimble is the right choice when the workload requires:

• Web search and retrieval for AI agents: complete, current, cited context from the public web

• Multi-step research, enrichment, or dataset building with Web Search Agents

• Structured outputs agents can use directly: schema-matched JSON, cited answers, clean markdown

• Enterprise governance: DPAs, audit trails, role-based access, Zero Data Retention on enterprise plans

• Delivery into production stacks: Snowflake, Databricks, S3, or BI tools

• Scraping known sites at scale, with Extract, Extraction Templates, or managed data services

Why these boundaries exist

• Legal and privacy risk reduction for customers and for site operators

• Protection of the open web through ethical collection and controls (including IP sourcing audits and site-communication headers)

• Enterprise-grade reliability: governance, validation, citations, and delivery into production stacks

Compliance and governance are core to Nimble's buyer value. For buyers whose workload falls inside these boundaries, the guardrails are a feature.

Examples of requests outside Nimble's scope

These fall outside the public-web-data boundary and cannot be supported:

• Logging into a private portal to export data

• Bypassing a paywall to retrieve subscription content

• Harvesting personal contact information (names, emails, phone numbers) from social profiles

• Copying an entire SaaS knowledge base that requires authentication

Get started

If the workload fits inside Nimble's scope:

• For AI agents: Nimble Platform, Products, and APIs and the MCP server spec

• For scraping and data feeds: Web Scraping with Nimble

• Pricing: Nimble Pricing