SignalMesh raises $40M to build a privacy-first search assistant for enterprise knowledge bases

Tech · 5 min read

SignalMesh announced a $40 million Series B round and the launch of its enterprise search assistant, which uses retrieval-augmented models with strict data access controls. The company says its architecture prevents model hallucinations by coupling answers with cited snippets and cryptographic attestations.

The assistant integrates with popular enterprise knowledge stores and offers role-based redaction and sticky policies so sensitive information is never surfaced to unauthorized queries. SignalMesh targets healthcare, legal, and finance customers that require both AI productivity gains and provable compliance.

Funding will expand integrations, runtime isolation features, and a partner program for SI firms. SignalMesh's roadmap includes offline indexing for air-gapped environments and native support for on-prem model inference.