Blumira Opens Hearth Pilot for Cross-Tool Security Investigations

Blumira opened a pilot on August 11 for Hearth, an AI security command center that investigates across an organization's existing security tools without requiring a Blumira platform subscription. The product queries connected systems and logs, returns findings with supporting evidence, and requires human approval before executing proposed actions.

Blumira Opens Hearth Pilot for Cross-Tool Security Investigations Blumira opened a pilot on August 11 for Hearth, an AI security command center that works across an organization's existing security tools rather than requiring a single-vendor stack. The product queries connected systems and logs, returns findings with supporting evidence, and keeps a human approval step before proposed actions are executed. Blumira opened a pilot on August 11 for Hearth , an AI security command center designed to investigate across an organization's existing security stack. The official product page invites organizations to join the pilot and says a Blumira platform subscription is not required. Hearth is positioned as a command layer over tools that security and IT teams already use. Blumira says it can connect to identity, endpoint, cloud, network, vulnerability-management, ticketing, source-control, password-management and other systems, while also using an organization's public attack surface. Existing Blumira customers can add the logs already parsed by the company's security operations platform. Evidence stays attached to the answer A user can ask a security question in plain language, review the investigation plan, and receive a finding that includes the evidence query and a recommended next step. The official page lists outputs such as ranked risk registers, compliance evidence, MITRE ATT&CK mappings, affected entities and scheduled security briefings. That evidence trail is the more consequential part of the launch. Generative interfaces can make security work faster, but an answer is difficult to trust or audit when the underlying query and source data are hidden. Hearth's stated design keeps those artifacts available for an analyst or auditor to inspect. The product also places a human checkpoint in front of actions and automations. Blumira says Hearth proposes an action and waits for approval before execution, which separates investigation assistance from autonomous remediation. A pilot, not a stack replacement SiliconANGLE reported that Hearth is intended for organizations using a mix of rival security products. Blumira's product page similarly describes it as vendor-agnostic and lists integrations across providers including CrowdStrike, Microsoft, SentinelOne, Splunk, Okta, AWS, Google Cloud, Cloudflare and GitHub. For security teams, the practical test will be whether the product can preserve consistent evidence and access controls across that breadth of systems. The official availability language is currently a pilot invitation, so claims about production performance, investigation speed and integration depth should be treated as vendor assertions until independently demonstrated in customer environments. Key Points - 1Hearth is offered as a pilot that can investigate across third-party security tools without requiring a Blumira platform subscription. - 2Blumira says findings retain their evidence queries, supporting analyst review, compliance work, and repeatable investigations. - 3Proposed actions require human approval, keeping the announced workflow short of autonomous remediation. Scoring Rationale Hearth is a relevant security-operations product that applies an AI interface across heterogeneous tools while retaining evidence and human approval. The launch is useful to security and data-platform practitioners, but the official availability is a pilot and the performance claims remain vendor assertions. Sources Primary source and supporting public references used for this report. Practice interview problems based on real data 1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with. Try 250 free problems /problems