Enterprise AI tools often fall short of being plug-and-play solutions. Without dedicated training on company specifics, agents struggle to grasp internal definitions such as revenue metrics or access controls. This challenge has prompted many AI providers to send engineers for custom integrations at customer sites.

Jedify, based in New York, targets this integration challenge directly. Its platform links to corporate knowledge repositories through APIs and constructs a context graph that supplies AI agents with relevant business details. Connected sources range from structured systems like databases, warehouses, lakes, SaaS applications, and BI tools to unstructured materials including reports, documentation, code repositories, Slack conversations, and meeting transcripts.

The company announced a $24 million Series A round led by Norwest. Returning investors S Capital VC and Cerca Partners joined the round alongside new backer Oceans Ventures. Snowflake participated strategically and is incorporating Jedify technology into offerings such as Cortex AI, Semantic Views, and CoWork.

Jedify emphasizes that effective enterprise agents require understanding of entity relationships, data flows, permission structures, domain expertise, workflows, operational norms, and proprietary terminology. This focused context enables agents to retrieve task-specific information without scanning entire corporate datasets.

Co-founder and CEO Assaf Henkin highlighted Kiteworks as a customer example. The compliance firm linked Snowflake, Tableau, Notion, and internal playbooks to the platform, then developed agentic applications supporting sales and account workflows. These tools generate on-demand briefings before client meetings and surface precise details during live discussions.

Henkin distinguishes the context graph from existing semantic layers, metadata catalogs, and knowledge graphs by noting its multi-dimensional nature. It tracks connections across entities, data assets, personnel, permissions, and clients while remaining model-agnostic and updating continuously with system changes.

Permission management forms a core capability. The platform inherits rules from identity providers, file systems, SaaS tools, and databases at row, column, and table levels. Customers can define additional groups to control agent access, supported by observability and governance features that monitor agent behavior.

Jedify focuses on mid-market and large enterprises with established data environments. It currently serves 10 to 20 early customers, including The Weather Company, and attracts interest from data-intensive industries such as gaming, industrials, and consumer packaged goods.

Snowflake's involvement stands out as major data platforms pursue similar features. Henkin maintains that Jedify complements these efforts because most organizational data and knowledge reside outside any single cloud environment. Building equivalent context layers internally can also prove expensive amid tightening scrutiny of AI token consumption.

As AI models advance and become more interchangeable, proprietary context layers may offer lasting competitive advantages. The new capital will support product development, team expansion, and go-to-market activities, bringing total funding to roughly $33 million.