Concentric AI

One-liner — DSPM built on deep-learning “Semantic Intelligence” that autonomously classifies data and fixes oversharing/permissions without writing rules or regexes.

What it does — Concentric’s pitch is autonomous data security posture management: rather than hand-built rules, regex, or labels, it uses deep-learning/NLP (“Semantic Intelligence”) to categorize structured and unstructured data, assess risk, and remediate — especially oversharing and risky permissions on documents and messages. It’s strongest on the unstructured-data problem (files in Microsoft 365/SharePoint/OneDrive, Google Drive, etc.): finding sensitive content, seeing who can access it, and fixing exposure. You buy it to cut oversharing risk before turning on AI assistants that would otherwise surface that data.

Where it sits in the stack — Primary: dspm; also data-access-governance. Data layer. Its risk role is controlling access to sensitive data — classify sensitive content and constrain who/what can access it (oversharing remediation), which directly limits what a Copilot/RAG system can surface. Trust zones: green/yellow unstructured-data plane.

Deployment & architecture — SaaS; connects to data repositories (M365/SharePoint/OneDrive, Google Workspace, file shares, cloud stores, messaging) to discover, classify, and remediate access. Agentless/API-connector model. Emphasis on autonomous classification + access remediation rather than appliance/agent deployment.

Positioning & differentiators — Differentiates on “no rules, no config” autonomous classification and on oversharing/access remediation for unstructured data — adjacent to what Varonis is known for, but ML-classification-led. Smaller and more focused than the category leaders: versus cyera/sentra (broad agentless cloud DSPM) Concentric leans into unstructured-content + permissions; versus bigid it’s lighter and more autonomous but narrower; versus microsoft-purview it spans beyond Microsoft and avoids label/rule heavy lifting.

Ownership, funding & M&AIndependent, privately held. Founded 2018 by Karthik Krishnan (CEO); HQ San Mateo, CA. Funding: $45M Series B (Oct 2024) led by Top Tier Capital Partners and HarbourVest Partners (Ballistic Ventures, Engineering Capital, Clear Ventures, Citi Ventures participating), total >$67M. No acquisition; no seed M&A flag. Ownership confidence: high.

CTO / hedge-fund lensDay-1 if your biggest data risk is unstructured-file oversharing in Microsoft 365/Google Workspace — exactly the exposure that lights up when you enable Copilot/RAG. Its autonomous-classification model means lower config burden than rule-based tools, attractive to a small security team. SR 11-7: supports AI-governance/data-exposure evidence indirectly. Caveat: smaller vendor (concentration/longevity risk); narrower than the cloud-datastore-broad leaders, so a multi-cloud-database-heavy fund may still need a cyera/bigid-class tool alongside it.

Competitors / alternativescyera, sentra, bigid, securiti, microsoft-purview, Varonis.

Open questions / to verify

  • Breadth of cloud/database (structured) coverage vs its unstructured-data strength.
  • Customer scale and any newer funding since the Oct 2024 Series B.

Sources

History

  • [2026-06-28] Stub created from seed registry.
  • [2026-06-28] Researched; established independent (>$67M raised; $45M Series B Oct 2024), founded 2018, San Mateo, autonomous “Semantic Intelligence” DSPM focused on unstructured-data oversharing/access remediation. Raised ownership_confidence to high. No M&A.