AI vendors sell the dream of autonomous agents that work out of the box. The reality is far messier. Most models fail to grasp how a specific company defines revenue, which employees can access sensitive files, or how internal teams actually collaborate. It is a fundamental disconnect.

New York-based startup Jedify is attacking this gap. The company announced today that it has raised $24 million in a Series A funding round led by Norwest. The goal is simple: give AI agents the institutional memory they currently lack.

Building a 'Context Graph'

Jedify connects to an enterprise’s existing data stack via APIs. It pulls from databases, data warehouses, SaaS applications, and even unstructured sources like Slack channels or meeting recordings. It then synthesizes this into what CEO Assaf Henkin calls a "context graph."

This isn't just another metadata catalog. It is a multi-dimensional map of relationships between people, data, permissions, and operational workflows. By mapping these connections, Jedify allows an AI agent to narrow its focus. Instead of searching the entire corporate digital footprint, the agent looks only at what is relevant to the task at hand.

"When you want to enable an agentic solution to really be autonomous, to drive decisions across CRM data, Zendesk tickets, and telemetry data, that’s when a context graph is much better than a semantic layer," Henkin said.

The Snowflake Connection

Data giant Snowflake participated in the round as a strategic investor. The partnership is telling. Snowflake is currently integrating Jedify’s technology into its own AI products, including Cortex AI and CoWork.

Some might view this as a threat to Jedify. If a platform like Snowflake builds its own context tools, why would a company need a third party? Henkin argues that the reality of enterprise data is fragmented. Most companies do not store all their knowledge in one cloud provider. They have multiple warehouses, various SaaS tools, and siloed databases.

"The big thing is that not all of your data is in those environments," Henkin noted. "Most of your knowledge is not there."

Security and Governance

Permissions remain the biggest hurdle for enterprise AI. An intern should not see the CFO’s revenue projections. Jedify addresses this by inheriting access rules directly from identity systems and databases.

It enforces row-, column-, and table-level access controls. Customers can also create custom groups to define exactly what an agent is allowed to touch. The platform includes observability tools, allowing IT teams to monitor agent behavior and ensure they are not hallucinating or leaking restricted information.

What This Means for Enterprises

Jedify is currently targeting mid-market and large enterprises with mature data stacks. With 10 to 20 early customers—including The Weather Company—the startup is proving that the demand for "context-aware" AI is real.

For companies trying to build this internally, the costs are often prohibitive. Training a model to understand the nuance of a business is expensive. It requires constant updates as information flows through the company. Jedify offers a shortcut. It keeps the context fresh in real time.

Key Takeaways

  • The Context Problem: AI agents struggle to perform because they lack specific knowledge of company workflows, permissions, and data relationships.
  • The Solution: Jedify’s "context graph" aggregates data from across the enterprise, creating a model-agnostic layer that agents can query for relevant, secure information.
  • Strategic Backing: Snowflake’s investment signals that major data platforms see Jedify as a necessary bridge for companies with fragmented, multi-cloud data environments.

What happens next depends on adoption. If Jedify can prove its graph scales across thousands of employees, it will become a critical piece of the enterprise AI stack. If it struggles with integration, it will remain a niche tool. The next six months of deployment will tell the story.