filtered.com

What It Is

Filtered Intelligence is not a traditional LMS or LXP, but a learning infrastructure layer for enterprise AI. It connects an organization’s existing learning content, skills data, LMS/LXP/HRIS systems, and content libraries to create a structured data layer that can be queried and called by AI agents in real time. The official site emphasizes “no new portal”: employees continue working in their existing tools, while Filtered sits underneath to provide content understanding, skills mapping, and AI access.

Core Capabilities

The product consists of four components: Ingest Content connects to and parses various learning assets, including SCORM packages; Map Skills uses AI to map content and employee data to skills frameworks; Signal Quality continuously evaluates content relevance, quality, and duplication; Connect AI uses MCP to let Copilot, Teams AI, Claude, ChatGPT Enterprise, and custom enterprise agents access a unified skills and content layer. Its key differentiator is an MCP and API-first architecture, aimed at turning enterprise learning assets from “invisible to AI” into “callable by AI.”

Pricing and Trial

The official site lists a starting price of around £70K/year, based on a predictable annual license fee. The final cost depends on the scale of content assets and deployment scope, rather than per-seat subscription pricing. Compared with the traditional LXP annual fees it cites at £500K–£1.5M+, its cost positioning is more of an enterprise-grade alternative. No free plan or self-service trial was found; only a 45-minute customized demo is available.

Pros and Cons

Its main strength is clear positioning: rather than replacing existing systems, it enhances the current learning technology stack through APIs and MCP. Its data security messaging is relatively comprehensive, claiming self-hosted AI processing, no use of content for public model training, and certifications including ISO 27001 and Cyber Essentials. Case studies mention results from AstraZeneca, GSK, ECITB, and others, such as learning content budget savings, faster curation, and reduced manual mapping costs. Limitations include limited disclosure around underlying models, Chinese-language capabilities, and quality evaluation methods. Deployment also depends on enterprise data governance, API permissions, and the maturity of existing systems, so it is not an out-of-the-box personal AI tool.

Who It’s For and Access from China

It is best suited to L&D, HR, enterprise AI, and IT architecture teams at large multinational companies, especially organizations with large content estates, complex skills frameworks, and a need for Copilot or enterprise agents to access learning data. It is less cost-effective for SMEs or teams that only need to generate course content or build a simple knowledge-base Q&A tool. The official site does not disclose China access, RMB payment, or localized support. Actual procurement may require evaluation of network connectivity, cross-border data compliance, and payment processes. Alternatives include traditional LXPs, an internally built enterprise skills data layer, or internal knowledge-base solutions based on the Microsoft ecosystem.