TechCrunch: Jedify raises $24M to arm AI agents with enterprise context

Jedify, a New York startup, announced a $24 million Series A on June 10, 2026, led by Norwest with participation from S Capital VC, Cerca Partners, Oceans Ventures, and Snowflake Ventures. The round brings total funding to approximately $33 million.

The company builds what it calls context graphs: a unified knowledge layer that connects AI agents to enterprise-specific information across databases, data warehouses, SaaS applications, and unstructured content such as documentation and Slack channels. The problem it addresses is common in enterprise AI deployments—general-purpose agents regularly fail not because of model quality but because they lack grounding in how a specific organization actually works. They don’t know how the company defines revenue, who has access to which file, or what the real workflows look like, and so they produce answers that are technically coherent but operationally wrong.

Context graphs differ from traditional semantic layers by capturing four dimensions at once: data relationships, roles and people, permissions and governance, and workflows. Crucially, the system inherits existing access controls from source systems, enforcing row-, column-, and table-level permissions across multiple connected tools. This addresses one of the harder compliance problems for enterprise AI deployments without requiring a separate permissioning layer. Early customers include The Weather Company.

For product managers working on enterprise AI features, the funding is a signal that “business context as infrastructure” is becoming its own product category. Teams that have run into the generic-output problem—where an AI agent gives answers that are correct in general but wrong for this company—are the market this solves for. The broader pattern is that enterprise AI deployment quality is shifting from model selection toward context layer quality, and tools like Jedify are emerging to serve that gap.