Jedify raises $24M to build a real-time context graph that connects enterprise data for AI agents
Jedify raises $24M Series A to build a real-time context graph linking enterprise data and permissions, enabling AI agents to act accurately across systems.
Jedify, a New York startup, has secured $24 million in a Series A round to commercialize a "context graph" designed to help AI agents operate inside complex enterprises. The company says its platform ingests data from databases, SaaS tools, BI systems and unstructured sources to create a multi-dimensional map of relationships, permissions and domain knowledge. Jedify’s approach aims to make AI agents more reliable and relevant by narrowing their focus to the exact data and rules needed for a task.
Jedify raises $24 million Series A led by Norwest
The Series A was led by Norwest with participation from returning investors S Capital VC and Cerca Partners, plus new backer Oceans Ventures, the company confirmed. Snowflake also participated as a strategic investor and plans to integrate Jedify’s technology with its AI offerings. The financing brings Jedify’s total capital raised to roughly $33 million and will be deployed toward product development, hiring and go-to-market expansion.
Context graph links data, permissions and workflows in real time
Jedify’s core product builds what it calls a context graph: a live representation of entities, data sources, people, permissions and operational assumptions. The graph is intended to be model-agnostic, updating in real time as information flows across connected systems so an AI agent can retrieve the most relevant context for a given task. The company positions this layer as distinct from traditional semantic layers, metadata catalogs or static knowledge graphs because it emphasizes relationships and live permissions over simple indexing.
Early customers test agentic workflows with integrated sources
Customers are using the platform to create agentic applications that combine dashboarding and real-time conversational capability. One cited example involved a compliance-focused customer linking Snowflake, Tableau and internal playbooks to surface tailored information for sales and account teams during customer interactions. Jedify says its early deployments — including pilots with organizations such as The Weather Company — show agents can proactively surface highly specific details, improving the quality of live conversations and operational workflows.
Snowflake takes strategic stake and product integration
Snowflake’s investment is framed as a strategic partnership rather than direct competition, according to Jedify’s leadership. The startup plans tight integration with Snowflake services, while arguing that most enterprises maintain data and knowledge across multiple clouds and systems. Jedify’s pitch is that it complements large data platform capabilities by unifying disparate sources and preserving company-specific context that larger vendors may not capture if data remains siloed.
Permissions and governance are built into the platform
A central challenge for agentic AI in enterprises is enforcing access controls so sensitive information does not leak to unauthorized users or workflows. Jedify says its context graph inherits and enforces identity and access rules from connected systems, including row-, column- and table-level controls. Customers can also define additional groups and policies to limit what agents and automated workflows can access, while observability and governance tools provide logs and controls to ensure agents behave as intended.
Target markets, technical differentiation and cost considerations
Jedify is targeting mid-market and large enterprises with mature data stacks and multiple warehouses or databases. The company sees particular interest from data-intensive sectors such as gaming, industrials and consumer packaged goods. Executives argue the platform can be less costly than training bespoke models to encode the same contextual relationships, especially as organizations increasingly manage AI token consumption and seek more efficient ways to apply large language models inside business workflows.
Jedify’s founders contend the platform’s durability will come from the proprietary context it captures rather than model exclusivity, since models continue to evolve and interchangeability increases. By focusing on entity relationships, permissions and workflow logic, they believe the context graph provides a repeatable layer that improves any downstream model’s performance.
The new funds will accelerate Jedify’s roadmap across product engineering, hiring and commercial work, the company said. As enterprises pilot agentic solutions, Jedify’s backers and partners will be watching whether a context-first approach becomes a practical standard for safely scaling AI inside complex organizations.
